Robin Sequence

Isolated and syndromic Robin sequence: dysregulation of the regulatory region upstream of SOX9 (17q24.3), Stickler collagenopathies (COL2A1, COL11A1, COL11A2), 22q11.2 deletion syndrome, Treacher Collins syndrome (TCOF1) and rare differential diagnoses with Robin sequence (TGDS, SNRPB, EIF4A3, SATB2, RBM10)

Literature and regulatory information updated: 10 October 2026.

0. Initial Reference Table

The following table gathers the genetic entities covered.

Sources of the second table: [5, 7, 12, 14, 15, 30–41].

1. Introduction

1.1. Founding articles and historical development

In 1923, Pierre Robin, a Paris stomatologist, described respiratory distress in infants with backward displacement of the base of the tongue (glossoptosis) and a small mandible, in an article entitled "La chute de la base de la langue considérée comme une nouvelle cause de gêne dans la respiration naso-pharyngienne"; in 1934 he added cleft palate to the picture, in an article published in the American Journal of Diseases of Children with the title "Glossoptosis due to atresia and hypotrophy of the mandible" [3]. For decades, the condition was called Pierre Robin syndrome, and the paediatric literature of the 1960s used that name, as in the work by Dennison published in Pediatrics in 1965; Latham studied the pathogenesis of the associated cleft palate in 1966 [1]. The first familial reports (siblings described by Smith and Stowe in 1961, by Sachtleben in 1964 and by Singh and colleagues in 1970, four siblings including twins by Shah and colleagues in 1970, and a family with transmission suggestive of dominant inheritance by Segreti and Maumenee in 1977) led OMIM to keep entry 261800 with autosomal recessive inheritance indicated; Opitz noted in 1973 that, especially in familial cases, Stickler syndrome should be the first hypothesis, and some of these siblings may have had Stickler or Wagner syndrome [6]. In 1975, Hanson and Smith described the U shape of the palatal defect and proposed that the cleft was the mechanical consequence of early mandibular hypoplasia, calling the condition "Robin anomalad"; the term "Robin sequence" was proposed in 1982 by Carey, Fineman and Ziter [1, 3].

1.2. The concept of sequence

In dysmorphology, a sequence is a pattern of anomalies in which one or more secondary morphological changes arise in cascade from a single malformation, disruption, dysplasia or deformation. RS is the canonical example: micrognathia prevents the tongue from descending, the retropositioned tongue lies between the palatal shelves and prevents their closure, and the same tongue, displaced into the oropharynx, obstructs the airway [3, 42]. This notion has practical consequences. RS is not an aetiological diagnosis but the description of a mechanism common to more than 170 OMIM entries; searching the phenotype term "Pierre-Robin sequence" in a facial analysis tool returns about 100 syndromes [3]. When RS coexists with a syndrome, the international consensus recommends naming both (for example, "RS with Stickler syndrome") [4].

1.3. Definitions and heterogeneity

The lack of a uniform definition has marked the RS literature. A 2015 European survey (101 responses, mainly from the United Kingdom, Germany, the Netherlands, Sweden and France) found 56 combinations of features used to define RS, with CP mandatory in 55%–96% of centres depending on the country and feeding difficulties required by 19%; in the Netherlands and Belgium, 14 teams used 6 different definitions, sometimes within the same centre; in the American Cleft Palate-Craniofacial Association, 73 responses gave 14 definitions, the combination of micrognathia, glossoptosis and CP being the most cited (44%) [3]. A review of 440 studies showed that the eponym "Pierre Robin sequence" was used in 62% and "Robin sequence" in 23%; the most frequent definition was micrognathia, glossoptosis and UAO, with or without CP (29%), and CP was required in 46% of articles, with a progressive shift, between 2009 and 2016, towards the definition without mandatory CP [3].

Reference books reflect this variation. The Oxford Desk Reference defines RS by the triad of micrognathia, CP and UAO, and notes that many clinicians use the term with only two of the three features [2]. The Brazilian textbook Síndromes Genéticas: Reconhecer para Tratar requires at least two of the features micrognathia, glossoptosis and cleft palate [10]. The 2016 Utrecht international consensus (145 participants from 24 countries) established that micrognathia is the primary feature, that glossoptosis and UAO are mandatory and that CP is a common, additional feature [4]. The 2024 European guideline, drawn up by an ERN CRANIO working group, adopted the three original criteria and added that, since there are no objective instruments for micrognathia and glossoptosis, the diagnosis rests mainly on OSA confirmed by a sleep study [3]. The 2021 French National Diagnosis and Care Protocol (PNDS), drawn up by the SPRATON reference centre, illustrates the persistence of this divergence. The centre endorsed the Utrecht consensus, but the PNDS adopts the definition still most used in France, the anatomical triad of retrognathia, glossoptosis and posterior median CP of any extent (velar or velopalatal, complete or incomplete), without requiring UAO, although this is frequent [18]. To make databases and epidemiological studies comparable, the same document recommends always recording whether it is "RS with cleft" or "RS without cleft" [18]. The PNDS attributes the 2010 survey that found 14 definitions (73 responses) to 204 specialists gathered at a Dutch cleft congress, whereas the European guideline places it among members of the American Cleft Palate-Craniofacial Association; the numbers coincide and the difference is only one of attribution [3, 18]. Throughout this article, the designation RS follows the 2016 and 2024 definition; when a study or guideline used another definition, this difference is indicated.

1.4. Aetiological classification and molecular milestones

The European guideline distinguishes clinically isolated RS, subdivided into confirmed isolated RS (up-to-date genetic testing without a diagnosis and a good clinical course at 18–24 months, apart from what results from RS itself) and presumed isolated RS (without up-to-date testing or a child younger than 18–24 months), from clinically non-isolated RS, subdivided into syndromic RS, with a genetically confirmed or only clinical diagnosis, and "RS plus", in which there are additional anomalies that do not constitute a recognisable syndrome [3]. In the absence of up-to-date testing, a genetic cause should continue to be considered [3]. The French PNDS uses the same logic with slightly shorter time frames: isolated RS is strongly suspected at birth in a term, normally grown newborn without other morphological, clinical or neurological anomaly, and is confirmed at about 18 months (between 12 and 18 months, in the genetic follow-up section) if all examinations are normal, with no myopia or radiological bone anomaly, and psychomotor development is normal; a patient with other anomalies and repeatedly negative genetic testing is classified as "RS plus" [18].

The molecular milestones of isolated RS are concentrated at the SOX9 locus. In 2004, Jamshidi and colleagues described isolated RS with a balanced translocation t(2;17); in 2007, Jakobsen and colleagues located the breakpoint of another t(2;17) 1.13 Mb upstream of SOX9 and proposed dysregulation of SOX9 and KCNJ2; in 2009, Benko and colleagues demonstrated, with linkage analysis, translocations, microdeletions and a point variant in a conserved element with enhancer activity, that some cases of RS result from disruption of very long-range regulatory elements of SOX9 [8, 43, 44]. In 2020, Long and colleagues identified the responsible enhancer clusters in human cranial neural crest cells and proposed that isolated RS is an "enhanceropathy" [45]. In parallel, the 2026 systematic review of the genetic landscape of RS showed that the monogenic contribution to isolated RS is limited and that SOX9 is the predominant locus [7].

1.5. Scope of the article

This article covers isolated and syndromic RS together, because the neonatal presentation, the respiratory and feeding assessment and much of the treatment are common. The European guideline formally applies to RS without a confirmed or strongly suspected syndrome and excludes RS with Stickler syndrome, craniofacial microsomia, Nager syndrome, Treacher Collins syndrome or 22q11.2 deletion, although it recognises that the functional approach overlaps [3]. Stickler syndromes, type II collagenopathies, campomelic dysplasia, 22q11.2 deletion syndrome and Treacher Collins syndrome are described only in the aspects relevant to the recognition and management of RS. Catel-Manzke (TGDS), cerebro-costo-mandibular (SNRPB), Richieri-Costa-Pereira (EIF4A3), SATB2-associated and TARP (RBM10) syndromes are treated as differential diagnoses with their own subsections.

2. Molecular Genetics and Pathophysiology

2.1. Mechanical pathogenesis and embryology

The classical model starts from hypoplasia of the mandibular region before 9 weeks of gestation. The small mandible keeps the tongue in a high, posterior position, between the vertical palatal shelves; these cannot become horizontal and meet in the midline, and the posterior shelves have to "grow over" the tongue. The result is a cleft of the secondary palate with a rounded, U-shaped outline, different from the inverted V shape of most primary cleft palates [1, 2]. The European guideline places the interference before the 10th week and describes the wide U-shaped cleft as typical, although the 2016 consensus stresses that the U shape is not constant [3, 4]. Normal palatal embryology frames the timing: the mesenchymal masses of the lateral palatine processes fuse with each other, with the nasal septum and with the median palatine process, and the upper lip and palate are complete at 12 weeks; in females, fusion of the shelves occurs about one week later [46]. Fetal magnetic resonance imaging (MRI) of fetuses later confirmed to have RS showed the tongue displaced into the cleft, rather than lying on the floor of the mouth, which supports mechanical interference [1, 3].

The purely mechanical model does not explain all cases. Electrophysiological studies suggest neuromuscular and brainstem dysfunction in isolated RS, an electromyographic study pointed to abnormal neuromuscular tone of the tongue, and the European guideline acknowledges that the exact pathophysiology is uncertain, with contributions from lingual and pharyngeal motor dysfunction and micrognathia of bony or neuromuscular origin [3, 4]. In the prospective Bauru series (HRAC-USP, Brazil), with 36 children with isolated RS followed up to 4 years, the cleft was complete and U-shaped in 75% and V-shaped in 25%, and the authors proposed that, in some cases, CP may be the primary event [47]. The review of ontogeny by Logjes and colleagues integrates these observations into a model in which RS results from oropharyngeal developmental events conditioned by a reduced stomodeal space, with an anatomical configuration that predisposes to CP [48]. The musculoskeletal components described include hyoid malformation, mandibular retroposition and reduced efficiency of the genioglossus in pulling the tongue forward [10].

The French PNDS places the process between the 45th and 55th days of embryonic life, the period of formation of the secondary palate: in the 7th week the mandible begins to grow downwards and forwards and draws the tongue with it, which allows the palatal shelves to approach and fuse, anteroposteriorly from the incisive foramen, by the 11th week [18]. The document organises the pathophysiology into three theories, now regarded as complementary and sometimes coexisting in the same patient. In the mechanical or anatomical theory, mandibular growth deficiency is primary, due to an anomaly of bone or connective tissue; this is the model of the mandibulofacial dysostoses and, according to the authors, the one that justifies early surgical approaches, in particular MDO. In the neurodevelopmental or functional theory, micrognathia is secondary to prenatal orofacial hypomobility, of central origin (brainstem, control of sucking-swallowing) or peripheral origin (myopathies, myasthenic syndromes), and the same neurological origin explains the micrognathia and the neonatal respiratory and feeding disorders; this model underpins non-invasive conservative approaches while awaiting mandibular growth. The third theory, of extrinsic compression of the mandible against the sternum (multiple pregnancy, oligohydramnios, uterine anomalies), is a minority view [18].

2.2. Aetiological categories and mandibular growth

Three groups of mechanisms produce the initial micrognathia: deformation due to restriction of fetal movement and growth (oligohydramnios, abnormal fetal position); neurological or neuromuscular deficit with poor fetal movement and akinesia, as in congenital myotonic dystrophy or Moebius sequence; and an intrinsic anomaly of the mandible or connective tissue, as in the collagenopathies and dysostoses [2, 49]. The distinction has prognostic consequences: in the deformation sequence, the mandible usually reaches normal size by about 2 years, whereas in syndromes with an intrinsic mandibular anomaly growth remains deficient; in Stickler syndrome, with concomitant maxillary hypoplasia, mandible and maxilla may become proportionate [2]. In the Dutch cohort of Basart and colleagues, RS due to deformation tended to improve spontaneously, whereas RS due to dysplasia (abnormal tissue in the mandible, larynx or pharynx) maintained abnormal postnatal development, which influences surgical airway management [3, 19]. This expectation of recovery is not confirmed in isolated RS studied long term: cephalometry shows a consistently shorter mandible, without complete catch-up growth (section 3.6) [3].

2.3. SOX9 and its upstream regulatory region

SOX9 (17q24.3; MANE transcript NM_000346.4) encodes a transcription factor with an HMG (high mobility group) box that acts as a regulator of chondrogenesis: it binds the ACAAAG motif in enhancers and super-enhancers and activates COL2A1, COL9A1, COL11A2, ACAN, SOX5 and SOX6; together with SF1 it activates anti-Müllerian hormone in testis determination [34, 40]. Regulation of COL2A1 by SOX9 explains the phenotypic overlap between campomelic dysplasia and spondyloepiphyseal dysplasia congenita [13]. ClinGen assigns SOX9 a haploinsufficiency score of 3, but classifies the evidence for the gene-disease relationship with isolated RS as "limited" [34].

The gene lies in a gene desert, within a topologically associating domain (TAD) that extends about 2 Mb upstream and contains tissue-specific enhancers. Variants in the coding region cause campomelic dysplasia with complete penetrance; translocations, inversions and deletions upstream leave the coding region intact, but reduce SOX9 expression in the tissues whose enhancer has been separated or removed [13, 50]. Severity decreases with distance from the gene: upstream breakpoints are distributed in three clusters, a proximal one at 50–375 kb (campomelic dysplasia, often with a difference of sex development in 46,XY individuals), a distal one at 789–932 kb, redefined by Fonseca and colleagues down to 585–601 kb (acampomelic campomelic dysplasia), and an "RS" cluster at 1.03–1.26 Mb (isolated RS); deletions of more than 1 Mb upstream are likewise associated with RS [9, 50]. Two individuals with breaks at 899 kb and 932 kb had skeletal findings so mild that the alteration was transmitted over several generations [13]. A common region of about 0.5 Mb upstream, when duplicated or deleted, causes an isolated difference of sex development, without dysplasia [13].

The studies that defined the isolated RS locus are as follows. Jakobsen and colleagues studied 10 Danish patients with non-syndromic RS and identified a t(2;17)(q23.3;q24.3) with a breakpoint 1.13 Mb upstream of SOX9 and 800 kb downstream of KCNJ2, with reduced expression of SOX9 and KCNJ2 measured by reverse transcription quantitative PCR (RT-qPCR) in five patients [44]. Benko and colleagues studied a four-generation family with 12 affected individuals and dominantly transmitted isolated RS, linked to a 5.4 Mb interval at 17q24.3-q25.1 (maximum lod score 3.32), in which a comparative genomic hybridisation microarray identified a heterozygous 75 kb deletion, located 1.38 Mb upstream of SOX9 and containing 10 highly conserved non-coding elements, which segregated with the disease; in 11 additional patients they found two de novo deletions (a centromeric one of more than 319 kb and a telomeric one of 36 kb) [6, 8]. The same authors found, in three families with autosomal dominant isolated RS, translocations with breakpoints at 17q24 clustered within 160 kb in a gene desert between KCNJ2 and SOX9 with more than 200 highly conserved non-coding elements, translocations clustered 1.06–1.23 Mb upstream, microdeletions of about 1.5 Mb centromeric and telomeric to SOX9 and a heterozygous point variant in a conserved element with enhancer activity that abolished function in vitro and altered MSX1 binding; in the developing mouse mandible, the chromatin of the 3 Mb region bounded by the deletions decompacted specifically in Sox9-expressing cells [8]. Gordon and colleagues described two new familial deletions and identified, by chromatin immunoprecipitation sequencing (ChIP-seq) for p300 in mouse craniofacial tissue, several enhancers validated in mouse and zebrafish, some within the deletions, and placed the RS locus at about 1.2–1.5 Mb [51]. Fukami and colleagues described a man with RS and mild hypoplasia of the left scapula, with a paracentric inversion 17q21.31–q24.3 and a microdeletion between −4.15 and −1.16 Mb, and concluded that the region beyond 1.16 Mb contains enhancers critical for the mandible [52]. Smyk and colleagues described a de novo deletion of about 1.58 Mb, which includes MAP2K6 and ends 1.28 Mb from SOX9, in a patient with RS and osteopenia with multiple fractures [53]. Castori and colleagues described a three-generation Italian family with a deletion of about 1 Mb including KCNJ2 and KCNJ16, dominant transmission of RS, skeletal signs of acampomelic campomelic dysplasia in some members and learning difficulties in the affected males, which demonstrates phenotypic continuity between the two entities and intrafamilial variability [11]. Amarillo and colleagues described an 18-month-old girl with isolated RS and a gene-free microdeletion of about 623 kb, inherited from her mother (who had CP), located about 725 kb upstream of the 5′ end of SOX9; this distance does not coincide with that of the clusters beyond 1 Mb and may reflect differences in genome build or reference point [54].

Long and colleagues, working with human cranial neural crest cells derived from stem cells, identified three enhancer clusters, EC1.45, EC1.35 and EC1.25, located 1.45, 1.35 and 1.25 Mb upstream of SOX9, active only in the cranial neural crest and fetal craniofacial tissue and inactive in chondrocytes [45]. Heterozygous deletion of EC1.45 reduced expression of the affected allele by 50%–55%; the effect of EC1.25 was smaller; EC1.45 activity depends on the "Coordinator" motif and on TWIST1 binding [45]. In the mouse, homozygous deletion of the EC1.45 orthologue reduced Sox9 expression by only about 13%, but reproducibly altered the shape of the mandible, and Sox9 haploinsufficiency in the neural crest produced micrognathia and CP with incomplete penetrance, which shows the particular sensitivity of the mandible to SOX9 dosage [45]. Patients' deletions rarely overlap and contain one cluster or the other; translocations that separate EC1.25 and EC1.45 from the gene appear to be associated with more severe phenotypes [45]. In 2025, a preclinical study showed that three variants of Neanderthal origin increase EC1.45 activity in zebrafish neural crest progenitors and that overexpression of SOX9 in these cells expands the cartilage precursors, which suggests that variation in this enhancer also modulates normal facial morphology [55].

From a diagnostic standpoint, these lesions (deletions in a gene desert, balanced translocations, point variants in enhancers and alterations of the TAD boundaries of the Sox9–Kcnj2 locus) are not detected by exome sequencing; deletions are detectable by microarray if probe density in the 17q24.3 region is sufficient, balanced translocations require karyotyping, long-read sequencing or genome sequencing, and point variants in enhancers can only be identified by genome or targeted sequencing [3, 7, 13]. These lesions are described by genomic coordinates with the genome build and by the distance to SOX9, because they have no coding description according to Human Genome Variation Society (HGVS) nomenclature; 48% of the studies included in the 2026 systematic review did not state the reference genome, which explains part of the discrepancies in distance between publications [7].

2.4. Other genes and loci in isolated RS

The systematic review by van de Velde and colleagues (107 studies, 777 subjects with RS, 244 with genetic results) listed, for clinically isolated RS, variants in BMPR1B, SLC39A11 (CNV), SOX9 (point variants and CNVs, in five studies), TBX22 and ZNF804B (CNV), 2p16 and Xp11 deletions, 14q and 22q11 duplications and a balanced translocation t(2;17) [7]. The authors concluded that the monogenic contribution to isolated RS is limited and that SOX9 predominates [7]. In the Melbourne series of 141 non-syndromic RS cases, sequencing of the conserved elements upstream of SOX9 in a subgroup of RS plus found only one substitution at a putative GATA1 binding site, inherited from an unaffected mother, which indicates that point variants in these elements are rare [56]. Outside the SOX9 locus, a family with RS was described (father, paternal aunt and proband with cleft of the secondary palate, micro-retrognathia, microstomia, malar hypoplasia and oligodontia) with a translocation t(1;2)(p34;q33) disrupting the first intron of FAF1, with reduced expression in the proband's lymphocytes; the FAF1 locus was also associated with CP in a study of 2530 trios with non-syndromic clefts [6]. The European guideline also mentions SLC10A7, identified by trio exome sequencing in a skeletal dysplasia with amelogenesis imperfecta and RS [3].

2.5. Collagens: Stickler syndrome and type II collagenopathies

Types II, IX and XI collagen are expressed in the vitreous, inner ear and cartilage; types II and XI form fibrils and type IX associates with their surface [57]. The α1(XI) chain, encoded by COL11A1, forms a heterotrimer with α2(XI) and α1(II) that nucleates and limits the diameter of type II collagen fibrils; splicing variants in the 54 bp exons of the C-terminal region are associated with the Marshall phenotype [36]. In COL2A1, most variants of the type II collagenopathies are glycine substitutions in the helical domain with a dominant-negative effect, whereas haploinsufficiency due to premature termination produces the Stickler phenotype with predominant ocular involvement [35]. The three dominant forms of Stickler syndrome (COL2A1, COL11A1 and COL11A2) may be associated with CP, RS, hearing loss and arthropathy; in the recessive forms due to COL9A1, COL9A2 and COL9A3 CP has not been described, whereas the recessive form due to biallelic COL11A1 includes profound deafness and CP, and LOXL3 has been described in two consanguineous families with high myopia, vitreous degeneration, CP and short stature [57]. Stickler syndrome is the most frequent genetic cause of RS [22, 57].

2.6. 22q11.2 deletion

22q11.2 deletion syndrome (OMIM 192430) is, in the 2026 systematic review, the most frequent chromosomal anomaly in non-isolated RS [7]. The palatal findings were quantified in a series of 1048 patients: CP in 28.5% (overt in 4.4% and submucous in 22.8%), cleft lip and palate in 0.6%, velopharyngeal incompetence in 55.2% and an overall palatal anomaly in 67% [14]. RS proper is occasional, in both the 22q11.2 deletion and duplication [58]. The proportion of 22q11.2 deletion among patients with RS varies between series (section 3.9).

2.7. Mandibulofacial dysostoses and disorders of the splicing machinery

Treacher Collins syndrome (TCOF1, POLR1C, POLR1D) presents with micrognathia, RS, CP and velopharyngeal insufficiency, and 60%–70% of cases result from de novo variants [2, 59]. Several syndromes with a craniofacial phenotype and RS result from spliceosome defects: mandibulofacial dysostosis Guion-Almeida type (EFTUD2), Nager syndrome (SF3B4), Richieri-Costa-Pereira syndrome (EIF4A3) and cerebro-costo-mandibular syndrome (SNRPB) [60]. In cerebro-costo-mandibular syndrome, SNRPB variants cluster in an alternative exon containing a premature termination codon ("poison" exon); the variants increase inclusion of this exon, the transcript is degraded by nonsense-mediated decay and the amount of SNRPB decreases without disappearing; no null variants or deletions have been found in survivors [38, 60]. In Richieri-Costa-Pereira syndrome, the mechanism is a biallelic repeat expansion in the 5′ untranslated region of EIF4A3, which also escapes exome sequencing [3, 7].

2.8. TGDS, SATB2 and RBM10

Catel-Manzke syndrome results from loss of TGDS activity. This enzyme produces UDP-4-keto-6-deoxyglucose, a metabolite that regenerates NAD+ in UXS1; the deficiency inactivates UXS1 and compromises the synthesis of skeletal glycans, according to work published in Nature in 2025 [37]. Inborn Metabolic Diseases classifies it among the defects of glycosaminoglycan synthesis, with CP, glossoptosis, micrognathia and bilateral hyperphalangy [61]. SATB2-associated syndrome results from intragenic variants, 2q33.1 deletions, translocations or duplications involving SATB2, usually de novo [15]. TARP syndrome, with X-linked recessive inheritance, results from hemizygous truncating variants (frameshift and nonsense) in RBM10 (Xp11.3), a gene expressed in the branchial arches and limbs of the mid-gestation mouse embryo, which is consistent with the phenotype; the molecular function of RBM10 is not described in detail in the sources consulted [41]. It accounted for 9 subjects with RS in the 2026 systematic review [7].

2.9. Chromosomal and teratogenic causes

In non-isolated RS, the 2026 review listed, in addition to the 22q11.2 deletion, deletions of 18q22, 1p36, 1q43q44, 2p16.3, 3q22.1, 4p16 (Wolf-Hirschhorn), 4q21–q34, 5p15 (cri-du-chat), 7q11.23 (Williams), 12q13, 16p11.2, 21q21–q22 and 22q12.2, duplications of 1q21, 15q11-q13 and 22pter-q12.3, balanced translocations involving 17q (t(8;17), t(17;18), t(17;20)) and mosaic trisomies [7]. Some findings (monosomy X, 16p11.2 deletion, 1q21.1 and 12p13.31 duplications) were considered probably unrelated to RS [7]. Terminal 4q deletion is associated with RS in 28% of cases and with 50% mortality before 15 months [62]. The 15q24 microdeletion includes RS among its occasional anomalies [63]. The teratogenic exposures described include alcohol, methotrexate, valproic acid and tamoxifen [10].

2.10. Gene-by-gene synthesis

Sources for the transcripts: VariantValidator, consulted in October 2026 [40]; mechanisms [7, 15, 35–38, 41, 45].

3. Clinical Spectrum and Manifestations by System

3.1. Manifestations by system

The triad of micrognathia, glossoptosis and UAO is accompanied by manifestations in other systems, which result from the sequence itself or from the underlying cause. The Brazilian textbook of genetic syndromes organises them as follows [10].

3.2. Upper airway obstruction

UAO has been described in 18% to 100% of patients with RS, depending on the definition and the assessment method; in the two population studies that quantified it, it reached 87%–93% [3, 17]. A healthy infant has fewer than one obstructive event per hour of sleep [3]. In RS, the obstruction is located at the base of the tongue, in the laryngopharynx or both, worsens in the supine position, during sleep and during feeding, and includes a spectrum of OSA, central apnoeas, hypoventilation and hypoxia; when awake, the infant may have stertor, recession, cyanotic episodes and brief resolved unexplained events (formerly called apparent life-threatening events) [3, 4]. Deep pectus excavatum with each inspiration, with suprasternal and intercostal recession, is frequent [2]. The obstruction may develop over the first 2 months: in a series of 11 infants, 10 had a respiratory problem, with late onset (day 24 to day 51) in 7 [1, 3]. Significant hypoxia may occur without obvious clinical signs [1]. Untreated obstruction is associated with cor pulmonale, faltering growth, metabolic, cardiovascular, neurocognitive and behavioural morbidity and death [1, 3].

Syndromic status does not predict the severity of OSA. In a study with PSG of at least 4 hours in 39 infants (15 with isolated RS), all had OSA, with no difference in severe OSA between isolated and non-isolated RS, and the need for intervention beyond prone positioning during admission was 53.3% in isolated and 79.2% in non-isolated RS (non-significant difference) [3]. In a Dutch cohort of 59 patients, with PSG in 42 at a median age of 47 days, there was likewise no difference between the groups [3]. In 71 infants younger than 90 days, the 49 who were not on respiratory support all had respiratory dysfunction on PSG (apnoea-hypopnoea index, AHI, below 10 per hour in 15, 10.1 to 20 in 15 and above 20 in 19); each additional millimetre of mandibular length on computed tomography (CT) reduced the risk of needing support by about 27% (odds ratio, OR, 0.73; 95% CI 0.56–0.96) [3]. Endoscopy showed additional airway anomalies in 17.4% of isolated RS and 55.6% of RS plus, and major difficulty in laryngeal exposure in 15.5% of 58 patients [3].

The natural history of OSA in the first year was described in a French prospective study of 45 infants (2018–2021), with PSG at 2 months and 8 months. Without treatment, in the supine position, OSA was severe in 77% at the first PSG; the median obstructive apnoea-hypopnoea index (OAHI) fell from 17 per hour to 5 per hour and sleep efficiency rose from 77% to 88% [64]. Under treatment at the first PSG, OSA remained severe in 81% of patients in the lateral position, 47% in the prone position and 85% on respiratory support, and PSG changed treatment in 40% (de-escalation in 18% and escalation in 22%) [64]. These data indicate spontaneous improvement to near-normal values at 8 months in most, but with persistent severe OSA in the first months despite positional treatment. The French PNDS describes a consistent time profile based on the reference centre's experience: in most children, the obstruction worsens over the first 4 weeks, remains stable and unsatisfactory for about 3 months and improves between 4 months and 1 year, with late re-worsening possible, especially in the syndromic forms; over the course of the day, it is greater in REM sleep and during feeds and less when calmly awake or with a non-nutritive dummy [18]. The obstruction has an anatomical component (collapse between the retropositioned tongue base and the posterior pharyngeal wall) and a functional component (hypotonia of the tongue base and pharyngolarynx and lack of coordination between swallowing and breathing) [18].

3.3. Feeding and growth

Feeding difficulties result from several mechanisms: anatomical alteration of sucking and swallowing, exhaustion and increased metabolic needs caused by the obstruction, reflux and regurgitation due to negative intrathoracic pressure, anomalies of the upper and lower oesophageal sphincters in 30%–70% and aspiration in 5%–65% [3]. The 2016 consensus proposes defining a feeding disorder in RS as the inability to ingest effectively oral quantities adequate for growth (for example, more than 30 minutes per bottle) or oral feeding that worsens the obstruction or causes aspiration; the difficulties may worsen in the first weeks and tend to improve from 4 months [4]. In a series of 24 isolated RS, 79% needed enteral feeding, for less than 3 weeks in 3, 3 weeks to 3 months in 2, 3 to 6 months in 4, 6 to 12 months in 9 and more than 12 months in 1; neither the severity of the micrognathia, glossoptosis or laryngomalacia nor syndromic status predicted prolonged enteral feeding [3]. Electromyography and manometry studies found oesophageal anomalies in 30%–54%, with reflux resistant to usual treatment and a tendency to spontaneous resolution after 12 months; in 20 isolated RS from Bauru, pH monitoring showed pathological reflux in 30% at 2 months, with a significant decrease at 4 and 6 months [3].

Growth is affected in the first year. In a study of 49 patients, 57% were malnourished (weight-for-age z score of −1 or less) at discharge and more than 30% had not been assessed by a dietitian; in 63 isolated RS, poor weight gain reached 27% and developmental delay 10%; in a cohort of 34 isolated RS, recovery occurred between 2 and 12 months and, at one year, most were within 1 standard deviation of the mean [3]. Poor weight gain may be the sign of an unrecognised chronic obstruction [3, 4].

The French PNDS adds the perspective of the Necker reference centre. The cleft, by itself, poorly explains the difficulties: infants with cleft lip, alveolus and palate, with a larger defect, have fewer sucking problems than infants with RS, and a recent study identified RS as the strongest risk factor for feeding difficulties among children with clefts [18]. The posterior position of the tongue is also insufficient, because non-nutritive sucking brings the tongue forward and immediately relieves the obstruction, the severity of glossoptosis did not correlate with that of the feeding difficulty and techniques that bring the tongue forward did not improve the ability to feed; the dominant mechanism would be a failure of sucking-swallowing-breathing coordination in the brainstem, supported by the electromyographic and oesophageal studies, to which are added the inhibition of sucking in any infant with respiratory difficulty and the increased energy expenditure due to recession [18]. The document distinguishes excessively long feeds (more than 30 minutes), reduced oral intake with insufficient growth, unsafe oral feeding (aspiration, episodes of distress) and respiratory worsening due to oral feeding; all grade 3 children need nutritional support, about two thirds of the children followed at the centre needed supportive enteral nutrition (EN), more prolonged in the syndromic and associated forms, and none was able to breastfeed [18]. In a prospective Necker series of functionally severe isolated RS (at least 3 months of EN), followed for 12 years, 74% no longer had feeding difficulties at 15 months, 18% still had them at 3 years and 6% at 6 years, a proportion lower than that observed in the general paediatric population; dysphagia for lumpy food, food phobias and intense gag reflexes are concentrated in the syndromic and associated forms, due to intellectual disability or a digestive motor disorder specific to the syndrome, as in Cornelia de Lange syndrome [18].

3.4. Cleft palate

CP is present in 80%–90% of patients; the typical cleft is a wide, U-shaped cleft of the secondary palate [3]. In the Bauru series, the cleft was complete and U-shaped in 75% and V-shaped in 25% [47]. RS without CP accounts for 8%–17% of cases, and epidemiological studies that included only RS with CP (68% of the studies in the 2023 meta-analysis) underestimate the prevalence [17]. The literature is contradictory as to whether the cleft is more extensive in RS than in isolated CP; one study found a cleft extending into the hard palate in 97% of RS and 54% of isolated CP [3, 4]. The 2016 consensus recommends describing the width, length, extent and shape of the cleft, using for example the Jensen classification (0, submucous cleft; 1, soft palate only; 2, soft palate and less than one third of the hard palate; 3, soft palate and between one and two thirds of the hard palate; 4, complete up to the incisive foramen) or the LAHSHAL code [4].

3.5. Hearing

Hearing loss is frequent, mainly conductive, mild to moderate and more common when there is CP; a Brazilian systematic review of 9 studies, with meta-analysis of 3, estimated for RS with CP an OR of 2.49 (95% CI 1.24–5.00) compared with controls, with low certainty [23]. Otitis media with effusion is frequent in any CP [46]. Ear anomalies, atresia of the external auditory canal and aplasia of the lateral semicircular canals may coexist, and point to syndromic causes [10].

3.6. Dentition and craniofacial growth

Tooth agenesis, excluding third molars, affects about 6.9% of the general population and just over 40% of patients with RS; two thirds of cases are bilateral and almost half involve both mandibular premolars [3]. A Bauru study described taurodontism in 92.73% [3]. Cephalometry of 96 isolated RS, compared with 50 isolated CP, showed a consistently shorter mandible at the three ages studied, with no change in the difference after 5 years; a three-dimensional study from Tübingen (37 RS younger than 8 years and 344 controls) found normal transverse and vertical growth with a constant sagittal deficit of the middle and lower thirds; another study of 34 isolated RS described a shorter cranial base, maxilla and mandible, more inclined planes and absence of catch-up growth [3]. Children with mandibular hypodontia had a smaller mandible, with a difference increasing with age [3]. In a series of 64 RS treated conservatively and followed beyond 13 years, 39.3% needed orthognathic surgery, with no difference between isolated and syndromic RS, with a predominance of class II and mandibular advancement more frequent than in isolated CP [3]. Prospective three-dimensional photography of 19 isolated RS treated with a palatal plate suggested some catch-up growth in the first year, although the mandible remained more retrognathic than in controls [65]. In syndromic RS, a systematic review of 202 cases from 20 syndromes found signs of balanced growth or apparent catch-up mainly in Stickler syndrome and 22q11.2 deletion, with limited evidence [66].

3.7. Speech and language

Speech studies in RS include only patients with CP and exclude syndromic RS [3]. Velopharyngeal dysfunction (VPD) ranged from 14.3% to 60% and secondary speech surgery from 13.4% to 47.6%; at 5 years, only 32% had articulation without cleft speech characteristics in one study [3]. Comparison with isolated CP shows contradictory results for VPD and secondary surgery, but 4 of 6 articulation studies found more frequent and more severe errors in RS [3]. The Swedish registry of 433 five-year-old children with cleft of the hard and soft palate confirmed this difference: in RS with additional conditions, the probability of adequate consonant production (OR 0.17; 95% CI 0.08–0.40) and of velopharyngeal competence (OR 0.24; 95% CI 0.10–0.56) was lower than in CP without additional conditions, and in RS without additional conditions adequate consonant production was also less likely (OR 0.50; 95% CI 0.29–0.88) [67]. The age at and technique of palatoplasty showed no consistent effect; the Furlow technique gave better nasality than the von Langenbeck technique in a Bauru study [3].

3.8. Neurodevelopment, sleep at school age and quality of life

A meta-analysis of 16 studies (1008 children) estimated a neurodevelopmental deficit after 1 year of age in 19% (95% CI 12–26), with 10% (5–16) in isolated RS, 19% (2–44) in syndromic RS, 63% (39–84) in RS plus and 35% (22–49) in non-isolated RS overall [68]. Prolonged neonatal hypoxia may be associated with developmental delay, but several studies found no significant cognitive or psychosocial differences in isolated RS [3]. In Tübingen, 21 school-age children treated with a palatal plate had a median OAHI of 1.3 per hour versus 0.4 per hour in controls, one non-syndromic child had severe OSA (OAHI 57 per hour) without complaints, and full-scale intelligence quotient was 102 versus 108 (p=0.05), within the normal range, with externalising behaviour in the at-risk range [69]. In Bauru, among 146 children aged 6 to 12 years with complete CP, OSA screening was positive in 59.42% of RS and 46.75% of isolated CP (non-significant difference), and home polygraphy compatible with mild to moderate OSA reached 89.65% and 78.94% [70]. In adolescence (72 patients aged 12 to 18 years, Necker), depression rates were favourable compared with the norm, oral quality of life was comparable to the norm and autonomy was lower; the only independent determinant of symptoms was vocal and oral quality of life [3]. The French PNDS, which cites the same Necker work (ADOROBIN study, then in press), stresses that the quality of life of adolescents was lower than that of controls in the dimensions of self-confidence, relationships with others and autonomy, and that the factors associated with worse quality of life in adolescents without intellectual disability were an associated disease (Stickler syndrome or equivalent) and oral, dental and buccal quality of life [18]. The same document describes, even in isolated RS, delayed gross motor milestones and, in some children, tense jaws, narrow mouth opening and poor facial expression, which justify a psychomotor assessment between 12 and 18 months [18].

3.9. Proportion of isolated and syndromic forms

The proportions vary with the definition of RS, the type of centre, the period, the length of follow-up and the extent of genetic testing. The following table gathers the main series.

Smith's textbook summarises the Dutch cohort as 38% RS in otherwise normal children and 62% syndromic, of whom 9% with a chromosomal anomaly, 29% with a Mendelian disorder (half Stickler) and 24% without a diagnosis, figures corresponding to those of Basart and colleagues [1, 19]. Shprintzen's series, cited by the Oxford Desk Reference, had fewer than 20% isolated cases, with Stickler in 34% and 22q11.2 deletion in 11% [2]. Reassessment changes the diagnosis: in the Dutch cohort, the initial diagnosis was changed in 48 of 191 patients (25.1%), mainly from isolated to non-isolated because of anomalies detected later; in the American series, 18% of isolated diagnoses became syndromic after more than 1 year and 4% moved in the opposite direction [3, 19]. In a follow-up of 64 infants published by Sheffield and colleagues in 1987, 12 of the 47 survivors (26%) received a syndrome diagnosis, most often Stickler syndrome (6 cases) [6]. In the Bauru series, 7 of 43 children initially classified as isolated RS (16%) later received a diagnosis of Stickler syndrome [47]. The French PNDS, based on the Necker series, summarises the proportions as about half isolated RS, 15% of them familial (10%–15% in the document's summary), and half forms with another malformation, more than two thirds of which fit an identified syndrome while the remaining third constitutes "RS plus"; advances in cytogenetics and in exome and genome sequencing are reducing this last fraction [18, 71]. In the 2026 systematic review, the most associated syndromes were 22q11.2 deletion, Stickler syndrome and Treacher Collins syndrome [7]. The frequency of the 22q11.2 deletion varies between series: 11% in Shprintzen's, part of the 33% of the non-isolated group in Nijmegen and 0.43% in Minneapolis, and the 2016 consensus considers it an uncommon cause of RS [2, 4, 20, 73].

3.10. Syndromic forms of particular relevance

Stickler syndrome. It is the most frequent syndrome in RS [22, 57]. The frequency of CP is 30%–60% with COL2A1, 60% with COL11A1 and 35% with COL11A2; myopia affects 80%–90% with COL2A1 and 80%–85% with COL11A1, and retinal detachment 40%–70% and less than 40%, respectively [12]. Micrognathia is most evident in infancy and becomes less distinctive with age; the cleft may be wide and U-shaped, narrow and V-shaped or submucous, and the latter often goes unnoticed [74]. In a survey of 316 patients, 95% had eye problems (retinal detachment 60%, myopia 90%, blindness 4%), 84% facial problems, 70% hearing loss and 90% joint problems [74]. In 31 patients with a molecular diagnosis, RS was present in 70.96% and 41.94% needed neonatal mandibular distraction osteogenesis (MDO) [75]; velopharyngeal insufficiency reached 16% [76]. The association of RS with myopia is highly suggestive of Stickler dysplasia [77].

Type II collagenopathies. Spondyloepiphyseal dysplasia congenita, Kniest dysplasia and achondrogenesis type II/hypochondrogenesis may present with RS, disproportionate short stature at birth and myopia, and imply cervical spine instability and anaesthetic risk [16, 35].

Campomelic and acampomelic campomelic dysplasia. The facies is similar to that of the type II collagenopathies, with RS and a flat midface; the long bones may be bowed, with pretibial dimples, and there are hypoplastic scapulae, 11 pairs of ribs and clubfeet; neonatal death results from airway instability (tracheobronchomalacia) or cervical spine instability, and about 75% of 46,XY individuals have ambiguous or female genitalia [13, 78]. Bowing is absent in about 10% (acampomelic form), which is overrepresented among upstream translocations and deletions [9, 13]. In the Bone Dysplasias atlas, an 18-month-old child with RS and small scapulae on palpation received a clinical diagnosis of the acampomelic form at 1.5 years, with straight femora (figures 20.1.1B and 20.1.5B of the atlas) [79].

22q11.2 deletion. Conotruncal heart disease, hypocalcaemia, immunodeficiency, velopharyngeal insufficiency with hypernasal speech, speech delay and feeding difficulties in about 36% of children accompany RS when it is present [2, 14]. Its detection requires investigation for cardiac, urogenital, skeletal, gastrointestinal and ophthalmological anomalies [3].

Treacher Collins syndrome. Downslanting palpebral fissures, coloboma of the lower eyelid, malar and zygomatic hypoplasia, bilateral microtia and atresia of the external auditory canal in one third, with normal intelligence [2].

Other syndromes. RS is occasional in Marshall syndrome, CHARGE syndrome, Kaufman oculocerebrofacial syndrome, boomerang dysplasia, lymphoedema-distichiasis syndrome (CP in 4%), Nager syndrome and mandibulofacial dysostosis with microcephaly [80–86]. In diastrophic dysplasia, two infants with a cleft and micrognathia similar to RS died of obstruction, and mortality from obstruction, including laryngeal stenosis, reaches 25% in early infancy [87]. In Moebius sequence, micrognathia of neuromuscular origin produces RS or bifid uvula in one third [49]. Yunis-Varón syndrome includes glossoptosis, and sternal malformation with vascular dysplasia includes occasional glossoptosis [88, 89]. RS has been described in chondrodysplasia with joint dislocations due to IMPAD1 and in Crane-Heise syndrome, which is lethal, with absent cervical vertebrae [90, 91]. The French PNDS, which counts more than 80 syndromes with RS, adds to the list of diagnoses most frequent in other series Loeys-Dietz syndrome, osteopathia striata with cranial sclerosis, atelosteogenesis, embryofetopathies due to alcohol, valproate and misoprostol, Kabuki syndrome and some congenital myopathies; among the genes most frequently implicated it cites COL2A1, the COL11 genes (COL11A1 and COL11A2), SOX9, KCNJ2, SATB2, SLC26A2, IMPAD1, EFTUD2, DLX5, DLX6 and EDN1 [18]. Catel-Manzke, cerebro-costo-mandibular, Richieri-Costa-Pereira, SATB2-associated and TARP syndromes are described in section 6.

3.11. Mortality: natural history and treated disease

Smith's textbook reports mortality of up to 30% in older series, much higher in syndromic RS [1]. More recent series of treated patients give lower values. In Utrecht, 10 of 103 consecutive patients (10%) died, at a median age of 0.8 years; 9 were syndromic, 7 died of respiratory failure (2 related to UAO), and the only isolated patient died of cerebral ischaemia after MDO; a neurological anomaly was associated with mortality of 40% versus 7% (OR 8.3; 95% CI 1.4–49.0), with no association with cardiac anomaly, neonatal airway surgery or prematurity; the literature reviewed by the authors gave mortality of 2% to 26% [24]. In Indianapolis, in a neonatal intensive care unit, mortality was 16.6% in 181 patients, associated with cardiac and central nervous system anomalies, with 2 deaths related to obstruction and no deaths in isolated RS [21]. In the American HCUP-KID administrative database, mortality was 6.6% (14.1% in syndromic and 2.2% in isolated RS) [25]. In the 2026 genetic systematic review, the 10 deaths recorded all occurred in non-isolated forms [7]. These differences reflect distinct populations (intensive care unit, reference centre, administrative database) and should not be read as the lifetime risk of an individual patient.

4. Clinical Recognition and Warning Signs

4.1. Prenatal period

Prenatal suspicion arises when the anomaly scan at about 20 weeks shows an abnormal facial profile with retrognathia or micrognathia; very severe micrognathia may be visible from 12–14 weeks, whereas isolated CP is difficult to detect even with three-dimensional ultrasound [2, 3, 46]. Polyhydramnios should raise the possibility of micrognathia severe enough to impair fetal swallowing [92]. Only 15%–30% of RS cases are diagnosed prenatally; in a survey of parents from eight European countries, RS had been suspected before birth in 19% [3, 93]. Prenatal suspicion is associated with a more severe postnatal phenotype: about 30% of cases identified in this way were Cole grade 3 and needed airway intervention [3]. Altered amniotic fluid volume (polyhydramnios or oligohydramnios) makes a syndromic diagnosis 2.3 times more likely (95% CI 1.3–6.3), with an OR of 4.18 for polyhydramnios (95% CI 1.6–10.9) [3]. The French PNDS documents the evolution of detection: in a Necker series published in 2001, with 117 consecutive cases, only 8 (7%) had been suspected on prenatal ultrasound, because of polyhydramnios, retrognathia, CP, vertical position of the tongue or a combination of these signs; since then, training of sonographers and the strict fetal profile, which in France has become a mandatory view of the anomaly scan at 22 weeks of gestation, have increased prenatal diagnosis [18, 71]. Posterior CP remains difficult to see: in a meta-analysis of 13 studies, the detection rate ranged from 0% to 22%; in data from 20 malformation registries in 12 European countries analysed in 2003 (709,030 fetuses), 198 had postnatal posterior CP (1 in 3581), 23 of whom had RS (1 in 30,827); and in a Norwegian prospective study (49,314 births, 1987–2004), none of the 24 posterior CPs (1 in 2054) had been seen on prenatal ultrasound [18]. The biometric criteria and referral are in section 5.8.

4.2. Newborn and infant

RS should be considered in any newborn with a small, recessed mandible, especially if there is CP, noisy breathing, recession, feeding difficulty or desaturations. An infant with respiratory compromise is examined without changing the position of the head and neck [2]. Clinical assessment includes observation in the supine, prone and lateral positions, during sleep, wakefulness and feeding, and the jaw thrust manoeuvre, which pulls the tongue forward and helps locate the obstruction at the tongue base [4]. Intraoral examination shows a poorly visible, posterior tongue, sometimes raised into the cleft [4]. Subjective clinical assessment is not sufficient: in a Tübingen crossover trial, assessment by professionals underestimated the obstructive apnoea index by a mean of 12.7 events per hour (95% CI 2.8–22.5) compared with polygraphy and correctly identified severe OSA in only 56% (supine) and 58% (prone), with no effect of profession or experience [94]. In a prospective study, patients with Cole grade 1 (no signs in the supine position) had OSA on PSG [3].

The French PNDS describes two frequent errors when the diagnosis is disclosed by a non-specialist in the maternity unit. In the first, RS is not recognised and a simple CP is discussed; these are usually moderate forms, but the feeding and respiratory difficulties are underestimated, failure of breastfeeding is not understood and the infant is discharged without the parents having been warned. In the second, the diagnosis is disclosed by someone without experience, with unjustified anxiety, isolated internet searching by the parents or urgent transfer not justified by the child's condition [18].

The revised Cole clinical classification distinguishes three grades: grade 1, no respiratory difficulty in the supine position; grade 2, mild, intermittent obstruction in the supine position, with some difficulty precipitated by feeding; grade 3, moderate to severe difficulty in the supine position and inability to feed orally [3]. The French PNDS presents it as the Couly classification revised by Cole and formulates it in terms of treatment: grade 1, minor disorders, facilitated autonomous sucking possible and tolerable obstruction treated by prone positioning; grade 2, moderate disorders, deficient sucking justifying EN and tolerable obstruction according to blood gas criteria that vary between teams; grade 3, significant obstruction requiring relief of the obstacle [18]. Some groups have defined grade 3 operationally as severe feeding difficulty and obstruction with arterial oxygen saturation (SaO2) below 90% for more than 5% of the time and partial pressure of CO2 (pCO2) above 50 mmHg and/or AHI above 10 per hour, or cyanotic episodes requiring intervention [3].

4.3. Alarm and emergency signs

Emergency signs are cyanotic episodes, apnoea, pallor, marked respiratory effort and SatO2 of 90% or less on continuous oximetry, which in the Brazilian protocol prompt nasopharyngeal intubation after nasofibroscopy [10]. An infant in severe acute respiratory distress is stabilised according to standardised paediatric resuscitation algorithms; a nasopharyngeal airway or laryngeal mask may stabilise the airway [3]. Deterioration may be late: respiratory difficulty may appear up to 2 months after birth, and a fall in weight or in the growth curve is a sign of deterioration [1, 3]. A normal capillary blood gas before day 9 of life may give false reassurance; pCO2 values of 55 mmHg and bicarbonate of 30 mEq/L have been used as cut-off points, and surveillance for CO2 retention should be maintained during the first 30 days [3]. The French PNDS defines the reasons for transferring a newborn with RS from the maternity unit to the neonatal unit of the regional university hospital: immediate or secondary respiratory difficulty, difficult to assess because obstructive hypopnoeas in the newborn easily go unnoticed (subtle recession) and the obstruction may suddenly cause a severe hypoxaemic episode; sucking-swallowing disorders that prevent feeding autonomy and discharge; and an associated malformative picture with suspicion of congenital heart disease. RS, even associated with other malformations, does not on its own justify immediate transfer, and precipitate transfers may traumatise the mother and the mother-infant bond [18]. Respiratory obstruction more severe than expected for the degree of micrognathia should prompt a search for multilevel obstruction (laryngomalacia, tracheobronchomalacia, tongue-base lesions) and skeletal causes such as the rib gaps of cerebro-costo-mandibular syndrome [3, 60, 95].

4.4. Clues to a syndromic cause

The search for an underlying cause begins at the initial examination and is repeated throughout follow-up. The findings that point to specific entities are summarised in the following table.

The history includes a three-generation pedigree with targeted enquiry about small chin, CP, eye problems, intellectual disability, infant deaths and osteoarthritis; the pregnancy history, with exposure to drugs (anticonvulsants, corticosteroids), alcohol and tobacco, maternal diabetes and amniotic fluid volume; neonatal feeding problems; and developmental milestones, vision and hearing [2, 4, 46]. Examination of the parents by a geneticist may identify microforms of clefting, bifid uvula, lip pits, micrognathia or other signs of a syndrome with variable expressivity [4, 46].

4.5. Newborn screening

There is no biochemical or genetic newborn screening for RS. Universal newborn hearing screening may, however, identify the hearing loss of the recessive forms of Stickler syndrome and of other associated syndromes [57].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criteria

The diagnosis of RS is clinical and functional. The 2016 consensus establishes that the diagnosis can be made in the presence of micrognathia, glossoptosis and UAO, and organises the assessment of each feature at two levels [4].

The 2024 European guideline recommends using the three original criteria (micrognathia, glossoptosis and airway obstruction), with the diagnosis based mainly on OSA confirmed by a sleep study, after screening by regular clinical assessment and continuous oximetry; it also recommends a baseline sleep study in all newborns with suspected RS and repeated studies to monitor treatment [3]. The guideline acknowledges that the quality of evidence is low (single-arm observational studies with high risk of bias) and that micrognathia and glossoptosis still lack objective classification [3]. Older criteria, still used in part of the literature and in textbooks, accept two of the three features or require CP [2, 10]. The French PNDS, which requires the triad of retrognathia, glossoptosis and CP (section 1.3), regards the diagnosis as a clinical observation of the newborn that does not require further investigations: retrognathia and glossoptosis are observed, and a median velar or velopalatal cleft and obstructive respiratory difficulties are sought; investigations serve to assess functional severity and aetiology, not to confirm the sequence [18].

5.2. Assessment of micrognathia

The diagnosis of micrognathia is largely subjective. Cephalometry is limited by two-dimensional imaging and by the difficulty of immobilising the infant; CT requires radiation and has no normative data [4]. The Jaw Index, proposed by the 2016 consensus, is calculated with a tape measure and a depth gauge (or a spatula): the tragus-to-tragus distance through the subnasal point (upper arc) and through the pogonion (lower arc) and the alveolar overjet are measured; the index is the overjet multiplied by the ratio of the upper arc to the lower arc [4]. In healthy newborns, interobserver agreement was 0.74 (CI 0.49–0.86) and intraobserver agreement 0.81 (CI 0.66–0.89); the index may not reflect functional airway compromise and needs validation in other populations [3, 4]. In 204 healthy 2-day-old newborns, measured with calipers, gonion-mandibular length was 5.1 ± 0.3 cm, gonion-maxillary length 5.4 ± 0.3 cm and the gonion-maxillary/gonion-mandibular ratio 1.06 ± 0.05, with a normal range of 0.96 to 1.16 [3]. Three-dimensional photogrammetry correlated with CT for mandibular volume (0.8799) and avoids radiation [3]. In 150 infants with isolated RS studied by CT, three morphological subtypes were described (short mandibular body; short body with obtuse angle; short body and ramus), with implications for osteotomy planning [3].

The French PNDS distinguishes retrognathia, which describes only the recessed position of the lower alveolar ridge relative to the upper one (in a healthy newborn at rest the two ridges are opposite each other), from micrognathia, which refers to the width of the mandible, and from micro-retrognathia, which combines the two; many studies use the three terms as synonyms [18]. For neonatal grading, it refers to the classification used by some authors, with the infant at rest, supine, measured with a small graduated caliper or by visual estimation: "moderate" retrognathia if the lower gum is less than 5 mm behind the upper, "medium" between 5 and 10 mm and "major" above 10 mm. It considers the Jaw Index the reference measure in the literature, but difficult to perform in practice, and facial imaging (radiography or CT) to quantify retrognathia a research tool [18].

5.3. Assessment of glossoptosis

Glossoptosis is a dynamic entity, defined as displacement of the tongue base into the oropharynx and hypopharynx; there is no defined correlation between the degree of micrognathia and that of glossoptosis [4]. Endoscopy is the minimum required to visualise the position of the tongue and identify other levels of narrowing; nasopharyngolaryngoscopy gives a more complete view than the transoral route, but an awake, agitated infant may not reproduce sleep conditions [4]. In 35 children with RS, awake endoscopy (median 5.5 days of life) had a sensitivity of 50.0% (95% CI 27.2–72.8) and a specificity of 86.7% (95% CI 59.5–98.3) for tongue-base obstruction compared with drug-induced sleep endoscopy (DISE, median 21 days), which detected significantly more obstructions [3]. The Yellon and de Sousa classifications were associated with clinical severity, but with sensitivities of 56.2% and 28.1% and low correlations with the Cole classification [3]. DISE carries anaesthetic risks [3].

The French reference centre proposed a three-stage clinical grading of glossoptosis, to correlate anatomical with functional data. The newborn is examined at rest, awake and calm, held by the examiner in a semi-sitting position at 45°; the mouth is gently opened and the mandible lowered with the thumb to observe the tongue [18].

Source of the table: [18]. The PNDS considers awake nasopharyngeal laryngoscopy often ineffective at locating the point of obstruction, because the infant is agitated and has excessive muscle tension, and, although it cites a study in which sleep endoscopy induced by sevoflurane and/or propofol had a minimal complication rate, it considers that, in practice, the risk of respiratory decompensation should not be taken; dynamic MRI and multidetector CT allow the obstruction to be quantified, but are not used in practice [18]. This position is more restrictive than that of the International Pediatric ORL Group consensus, which proposes DISE for unexplained obstruction before or after surgery, and contrasts with the greater sensitivity of DISE described above [3, 26].

5.4. Respiratory assessment

Sleep study. PSG or polygraphy for at least 4 to 6 hours, with at least 3 hours of sleep, allows the infant to be assessed; polygraphy includes body position, actigraphy, nasal flow, thoracic and abdominal belts, oxygen saturation, transcutaneous CO2 and video or nursing observation, and event scoring follows the American Academy of Sleep Medicine [4]. The European guideline considers cardiorespiratory PSG, without electroencephalography, sufficient for all infants with suspected RS, to be performed or repeated at 2–3 weeks of life or later, because of neonatal immaturity [3]. The 2016 consensus proposed a three-level approach: with no respiratory signs observed by an experienced clinician for an adequate time, PSG may not be mandatory; with any sign, referral to a centre with sleep studies; with severe obstruction, urgent treatment, even without PSG [4]. The Tübingen group advocates cardiorespiratory polygraphy in most patients and disputes systematic full PSG [93]. The French PNDS adopts the same position: for the coordinating centre, respiratory polygraphy without electroencephalography suffices, because the aim is to detect and grade OSA and any central apnoeas, and knowing in which sleep stage they occur does not change management; in a Lyon study of 20 children undergoing full PSG, blinded reanalysis without electroencephalography gave similar AHIs, with higher obstructive apnoea indices and lower obstructive hypopnoea indices that offset each other, and the scarcity of PSG laboratories in France makes this finding important [18]. When only SpO2 and transcutaneous CO2 pressure (PtcCO2) recording is available, the PNDS considers it a screening test: pathological results (mean PtcCO2 above 50 mmHg or more than 5% of the time with SaO2 below 90%) require treatment, normal results with a perfect clinical picture allow close monitoring, and all intermediate situations require polygraphy or PSG [18].

Clinical assessment of severity. The PNDS lists the clinical elements on which teams agree for assessing the severity of the obstruction: the height and weight growth curve, the calories needed to achieve normal growth, comfort and quality of wakefulness (often in the prone position), psychomotor development, calm and restorative sleep and the absence of cyanotic episodes and episodes of distress [18].

Thresholds. There are no universally accepted criteria. The European Respiratory Society statement for children aged 1 to 23 months classifies OSA as mild (AHI 1–5 per hour), moderate (more than 5 to 10) and severe (more than 10), and recommends treatment with an AHI above 5, or 1 to 5 with symptoms or risk factors [3]. The OAHI is preferable to the total AHI, because central apnoeas vary with age; some groups use the mixed and obstructive apnoea index (MOAI), with greater interobserver agreement [3]. The following table shows the diversity of thresholds published in RS [4].

Oximetry and blood gases. Oximetry is specific but insensitive for OSA: in 122 isolated RS, the number of desaturations below 80% per hour could be normal with moderate to severe OSA on PSG [3, 4]. Peripheral oxygen saturation (SpO2) below 90% for more than 5% of the time was considered significant in 62 isolated RS, without formal justification of the threshold; up to 5% of healthy infants spend more than 4% of the time with SpO2 at or below 90% [3]. Oximetry serves for screening and home monitoring, but does not diagnose OSA [3]. In 274 patients from two centres, infants who needed intervention had higher capillary pCO2 and bicarbonate on days 0–6 and 7–30, with cut-off points of 55 mmHg and 30 mEq/L; a normal blood gas after day 9 makes intervention less likely [3].

Flexible nasendoscopy. FNE confirms tongue-base obstruction and identifies additional levels (laryngomalacia, subglottic anomalies); in 52 patients it had a sensitivity of 87% and a specificity of 100% for OSA confirmed by PSG, with rare complications; in a series of 60 isolated RS without sedation, it was abnormal in 45 (75%) [3]. The Sher endoscopic classification, applied in the first month and at 6 and 12 months in 56 patients, did not correlate with clinical severity, but types 1 and 2 progressed to no symptoms or type 1 at 12 months, whereas types 3 and 4 persisted throughout the first year [3]. The European guideline recommends FNE in all patients with RS, with sedation decided case by case, and recording of intubation difficulties [3]. Before any surgery, endoscopy of the whole airway is needed: in 83 infants with micrognathia and obstruction studied by laryngobronchoscopy, 23 (28%) had tracheal or laryngeal problems that made MDO unfeasible [3].

The French PNDS describes nasofibroscopy as the examination that specifies the level of obstruction (glossal, glossopharyngeal or glossopharyngolaryngeal) and identifies some of the associated anatomical anomalies, and considers complete collapse of the tongue base against the pharyngeal wall, an omega-shaped or tubular epiglottis and anteroposterior laryngomalacia to be signs of severe glossopharyngeal hypotonia [18]. When the infant is awake, the examination underestimates the picture because the infant cries; under general anaesthesia, it risks respiratory deterioration and requires a sound clinical rationale; some propose light, transient sedation; and signs of a lower obstruction may indicate full endoscopy under general anaesthesia, with great perioperative precautions [18]. In addition to the Sher classification (1986, 4 types of increasing severity, initially in obstructed children without RS) and a Mexican grading correlated with PSG, the PNDS proposes that otorhinolaryngologists use a common 12-item grid, each item scored from 1 to 4 [18].

Source of the table: [18]. Helical CT allows measurement of airway calibres, the mandibular body and ramus and the distance between the pharyngeal wall and the hyoid, with three-dimensional reconstruction; some US teams use it in the decision algorithm, but in France it is a research technique [18].

Criteria for relieving the obstruction. According to the PNDS, an obstruction is not clinically tolerable when it compromises growth, causes episodes of distress or cyanotic episodes, or compromises development, sleep or comfort [18]. Blood gas criteria are not agreed upon: a review by the international Robin consensus group identified 32 methods for assessing obstruction, and the most cited criteria are SpO2 below 95% for more than 5% of a 24-hour recording, PtcCO2 above 50 mmHg, an obstructive or mixed apnoea-hypopnoea index above 10 per hour and a desaturation index above 3% [18]. These thresholds differ from the OAHI above 5 per hour used as guidance by the European guideline (section 7.3) [3].

5.5. Feeding assessment

The 2016 consensus recommends feeding assessment by experienced professionals in the first days and weeks, including the quality and duration of bottle feeding, the effect of feeding on respiratory status and the quality of weight gain (expressed in standard deviations) [4]. Swallowing endoscopy, videofluoroscopy, electromyography and oesophageal manometry have been described, but there is no consensus on their validity and interpretation in infants with RS; the European guideline reserves them for complex cases that do not respond to treatment [3, 4]. The French PNDS considers none of these examinations indispensable: fibreoptic swallowing evaluation is not needed to start tube EN, but may help when gastrostomy is being discussed in non-isolated RS with a neurological deficit in which aspiration cannot be assessed clinically; swallowing electromyography (which showed lack of coordination between the submental muscles, innervated by cranial nerve IX, and the muscles innervated by cranial nerve X) and oesophageal manometry are no longer used routinely, because the clinical practice of an experienced team suffices to decide whether to continue bottle feeding; and suckometry is a research tool (DYSROBIN study) [18].

5.6. Aetiological assessment

All infants with RS should be assessed by a paediatrician and a clinical geneticist, with reassessment, because the syndrome may become apparent later [4]. The aetiological assessment recommended by the 2016 consensus includes the following elements [4].

The consensus also recommends an echocardiogram when 22q11.2 deletion is suspected or there is a murmur, brain MRI if there is abnormal tone, fetal akinesia, arthrogryposis or microcephaly, and skeletal radiographs if dysplasia is suspected; radiographs of the scapula, ribs and long bones help identify lesions around SOX9 [4]. The Indianapolis series, in which mortality was associated with cardiac and central nervous system anomalies, proposed echocardiography and cranial imaging in the initial assessment [21]. A child with RS should have the index fingers examined and radiographed (Catel-Manzke), and a chest radiograph should look for rib gaps when the obstruction is more severe than expected [2, 37, 60]. Slit-lamp examination distinguishes the membranous vitreous (type 1, COL2A1) from the beaded vitreous (type 2, COL11A1) [36]. Measurement of 7-dehydrocholesterol is indicated if there are other signs of Smith-Lemli-Opitz syndrome [46].

The European guideline makes the following genetic recommendations: refer all newborns and infants with RS as early as possible to a clinical geneticist at a reference centre, or at 6 months if early referral is not possible; re-refer between 2 and 3 years all children without a diagnosis after the initial work-up, or earlier if new findings appear; perform genetic testing including, at a minimum, CNV analysis and a panel of RS-associated genes on exome data (preferably as a trio); where this is not possible, CNV analysis and a Stickler syndrome-targeted panel [3]. The guideline acknowledges that, of the 14 studies read in full text, none allowed estimation of the overall yield or test accuracy in RS, so the recommendations are based on expert opinion [3]. The 2026 systematic review also concluded that it is impossible to estimate diagnostic yield separately in isolated and non-isolated RS [7]. The 2016 consensus stated that clinical exome sequencing provided a diagnosis in up to one third of individuals without a diagnosis by other methods, in the context of rare diseases in general and not specifically in RS [4].

Sources of the table: [3, 7, 13, 38]. In cerebro-costo-mandibular syndrome, the alternative exon of SNRPB is often missing from standard annotations, and exomes or panels filtered for annotated coding exons may fail to detect the variant; in a 2023 genome study, the variant was only found after adding the annotations of the "poison" exons, so the request should confirm coverage of this exon [38]. Unsolicited findings in exome or genome sequencing occur in less than 1% [3]. A negative result does not exclude a genetic cause, because a large proportion of the known lesions of isolated RS escape first-line tests [3, 7].

The French PNDS organises the aetiological assessment of the newborn into two scenarios, according to the initial examination [18].

Source of the table: [18]. If the initial hypothesis was isolated RS and the minimum assessment excluded a collagenopathy, but morphological or, above all, neurological anomalies appear in the first year, the assessment is extended to the second scenario, with MRI, genetics clinic, chromosomal microarray analysis (CGH array) and a molecular study targeting the genes for cleft palate and intellectual disability (SATB2 in particular) or for muscle [18]. Genetic testing starts with CGH array, unless the family history or clinical picture points immediately to a diagnosis, and continues with panels of genes for mandibular development, growth or the branchial arches, whose composition varies between teams; in isolated RS, the PNDS considers molecular analysis currently of little value, because only a few variants in or upstream of SOX9 have been described, and in associated forms without a diagnosis it provides for continuation with exome or genome sequencing and reassessment by a specialised genetics team every 2 or 3 years [18]. The strategy differs from the European guideline, which proposes from the outset CNV analysis and a trio exome panel for all and a single scheduled reassessment at 2–3 years [3].

5.7. Report and nomenclature

The report of a point variant should state the gene, the MANE transcript with version, the coding and protein HGVS description, the zygosity and the class according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP); for rearrangements upstream of SOX9, it should state the genomic coordinates with the genome build, the size and the distance to the gene, and whether the alteration is de novo or inherited [7]. The MANE Select transcripts confirmed in VariantValidator are SOX9 NM_000346.4, COL2A1 NM_001844.5, COL11A1 NM_001854.4, TCOF1 NM_001371623.1, SNRPB NM_003091.4, TGDS NM_014305.4, SATB2 NM_001172509.2 and RBM10 NM_005676.5 [40]. The most frequent TGDS variant in Catel-Manzke syndrome is normalised as NM_014305.4:c.298G>T, NP_055120.1:p.(Ala100Ser), NC_000013.11:g.94590868C>A on GRCh38; in ClinVar (VCV000162455) it is classified as pathogenic or likely pathogenic, with several submitters and no conflicts, last evaluated in December 2025 [40, 97]. The breakpoints of the case described in section 10 were published on hg19 (chr17:69,516,104-69,531,685) [9]. Variants of uncertain significance require segregation studies and periodic reassessment, which the guideline and the 2026 review also recommend for cases without a diagnosis [3, 7].

5.8. Prenatal diagnosis

A definitive diagnosis of RS is only possible after birth, because it depends on the obstruction. Fetal ultrasound and MRI identify risk markers [3].

The European recommendations are: consider the diagnosis if routine ultrasound shows an abnormal facial profile; refer to a reference centre (additional ultrasound and/or MRI) when there is suspicion or a family history of micrognathia; at the reference centre, consider RS with an inferior facial angle below 45.5° or mandibular length 2 or more standard deviations below the mean from 18 weeks; with MRI, consider RS in the presence of isolated CP, a tongue shape index above 80% and an inferior facial angle below 48°; offer genetic counselling; provide information on respiratory and feeding difficulties; and plan delivery with a doctor experienced in the difficult neonatal airway present [3]. Genetic counselling is useful when retrognathia is detected before 21 weeks, and after 24 weeks it mainly informs peripartum decisions [3]. The quality of evidence is low, with overlapping populations from two centres (Necker and Boston) [3].

The available prenatal genetic studies refer to fetal malformations in general and not specifically to RS: aneuploidy or CNVs explain up to 40% of pregnancies with malformations; trio exome sequencing gave an additional yield of 8.5% (160 trios) and 10.3% (234 trios), Dutch rapid exome sequencing 33% (18/54), with an impact on decisions in 68%, and prenatal genome sequencing 19.8% (22/110) [3]. SOX9 variants may give isolated RS or campomelic dysplasia, with very different prognoses, which complicates counselling even when the variant is identified [7]. In Stickler syndrome, the 19–20-week scan may detect CP, but its absence does not exclude the diagnosis [12].

The French PNDS describes the sonographic signs to look for and the investigation pathway. Posterior CP should be suspected when the tip of the tongue appears to be above the plane of the palate; three-dimensional analysis of the palate and two techniques for visualising the uvula (the "equals sign") have been proposed, and the indirect signs that alert the sonographer are retrognathia, glossoptosis and polyhydramnios [18]. Retrognathia is measured on a strictly sagittal view, with an inferior facial angle below 50°; it is not very specific, may be present in healthy fetuses and also occurs in chromosomal anomalies and in multiple bone syndromes (acrofacial and mandibulofacial dysostoses, oral-facial-digital syndromes, cerebro-costo-mandibular syndrome, chondrodysplasias, Goldenhar syndrome) or neurological syndromes, so it should prompt a search for the other elements of the triad and, above all, any other malformation and any sign of fetal hypomobility [18]. On the fetal MRI of Resnick and colleagues (2018), the tongue was shorter and the shape index greater in RS than in fetuses with micrognathia without RS, and the tongue base touched the posterior pharyngeal wall in 19% of RS cases, a finding absent in the comparison groups [18]. Agreement between prenatal suspicion and postnatal diagnosis is very good in fetuses with a family history of RS or Stickler syndrome and good when there are other malformations, but, in the Necker experience from 2002 to 2016, 22% of prenatal suspicions of isolated RS turned out to be non-isolated RS after birth [18].

The PNDS prenatal algorithm has three steps: refer the parents to a reference sonographer to confirm the suspicion and look for associated signs, with attention to the central nervous system, face, bones, extremities and fetal mobility; use fetal MRI systematically or only in case of doubt, depending on local experience and the parents' wishes; and present the case to a multidisciplinary prenatal diagnosis centre (CPDPN), with a minimum assessment including a genetics clinic appointment and CGH array on amniotic fluid, plus electrophoresis of sterols in amniotic fluid, looking for Smith-Lemli-Opitz syndrome, if there is intrauterine growth restriction or a suggestive malformative association (genitalia, microcephaly, heart disease, renal, cerebral or limb anomalies) [18].

5.9. Diagnostic algorithm

The French PNDS adds to this pathway, for probably isolated RS, neonatal serum calcium, ophthalmological examination with retinoscopy at about 6 months and skeletal radiographs at about 1 year; in the prenatal period, presentation to the CPDPN with CGH array on amniotic fluid; and, for "RS plus", genetic reassessment every 2 or 3 years [18].

6. Differential Diagnosis

The differential diagnosis of RS has two levels. The first is functional: distinguishing tongue-base obstruction from other causes of neonatal respiratory difficulty (choanal atresia or stenosis, laryngomalacia, tracheobronchomalacia, tongue-base lesions, central hypoventilation), which may coexist with RS and require endoscopy of the whole airway [3, 56, 95]. The second is aetiological: identifying the syndrome or sequence that produced the micrognathia. OMIM lists 483 genetic diseases with micrognathia and POSSUMweb 1080 genetic and teratogenic diseases [4].

6.1. Comparative table

Sources of the table: [2, 7, 13–16, 32, 35, 38, 39, 49, 57, 60, 62, 79–81, 87, 90, 96, 98].

6.2. Catel-Manzke syndrome (TGDS)

Catel-Manzke syndrome (ORPHA:1388; OMIM 616145) is autosomal recessive, with an estimated prevalence below 1 per 1,000,000 and about 33 cases described [30, 39]. Most present with RS and growth delay due to feeding and respiratory difficulties, and may need surgery; CP is present in 78%, micrognathia in 72%, malformed ears in 33%, hyperphalangy of the index finger (with radial deviation and ulnar clinodactyly) in 100% of the patients in the series of Smith's textbook, clinodactyly of the 5th finger in 39%, single palmar crease in 40%, heart disease (septal defects, overriding aorta, coarctation, dextrocardia, ductus arteriosus) in 39% and postnatal growth deficiency in 75%; the great majority of survivors have normal intelligence [39]. Prenatally, nuchal oedema, heart disease, clubfeet and retrognathia have been described [39]. Hand radiography shows an accessory ossicle at the base of the index finger [37]. Absence of hyperphalangy does not exclude the diagnosis: a 12-month-old boy with RS without CP and with hyperphalangy was at the severe end, and two individuals with biallelic TGDS variants had no radiological hyperphalangy but had disproportionate short stature [37]. The variant c.298G>T p.(Ala100Ser) is present on at least one allele in most patients described and is considered by Smith's textbook a probable European founder variant [39, 97]. Radiological differential diagnoses include Desbuquois dysplasia, IMPAD1 chondrodysplasia, Chitayat syndrome (ERF), Temtamy syndrome (CHSY1), brachydactyly type C and biallelic KYNU deficiency [37, 99]. There is no biochemical test of TGDS activity in clinical use [37].

6.3. Cerebro-costo-mandibular syndrome (SNRPB)

Cerebro-costo-mandibular syndrome (ORPHA:1393; OMIM 117650) presents with severe micrognathia and glossoptosis, a short or U-shaped cleft of the soft palate, a small bell-shaped chest and gaps between the ossified posterior and cartilaginous anterior portions of the ribs (from the 4th to the 10th, especially the 5th), with absence of some ribs (12th, 1st, 11th) [31, 60]. Neonatal manifestations are variable and may be limited to RS, so the diagnosis should be considered in any newborn with RS and respiratory problems more severe than expected; a chest radiograph confirms it [60]. In the British series of 16 patients, severe micrognathia and rib gaps were constant, and CP, feeding difficulties, neonatal respiratory distress, tracheostomy and scoliosis were common; 12 of 16 cases were sporadic and an SNRPB variant was identified in 12 of 14 tested; the number of absent or vestigial ribs, more than the number of gaps, predicts respiratory difficulty and lethality [38]. Microcephaly affects 25%, scoliosis half of the survivors, and the intellectual disability of severe cases is probably secondary to neonatal hypoxia [60]. A similar phenotype, with rib fusions, has been described in siblings with a COG1 variant [100].

6.4. Richieri-Costa-Pereira syndrome (EIF4A3)

Richieri-Costa-Pereira syndrome (ORPHA:3102; OMIM 268305) is autosomal recessive, with about 33 cases described, and combines RS, median cleft of the mandible and radial and ulnar ray defects, including fusion of metacarpals IV and V [32, 96]. The mechanism is a biallelic repeat expansion in the 5′ untranslated region of EIF4A3, which exome sequencing does not detect [3, 7].

6.5. SATB2-associated syndrome

In SATB2-associated syndrome, CP affects 45%, micrognathia 42% (rarely with surgical correction) and palatal anomalies overall 76% (cleft, high-arched palate, velopharyngeal insufficiency, bifid uvula in 3%) [15]. All patients have a neurodevelopmental deficit, most with absent or almost absent speech, normal growth, drooling and dental anomalies [2]. In the 2026 systematic review, SATB2 was the third most frequent gene in non-isolated RS (13 subjects), also in presentations classified as RS plus [7]. In the neonatal period, when developmental delay is not yet evident, the syndrome may be mistaken for isolated RS, which reinforces reassessment between 2 and 3 years [3].

6.6. TARP syndrome (RBM10) and other genes in non-isolated RS

TARP syndrome is named after its four classical features: talipes equinovarus, atrial septal defect, RS and persistent left superior vena cava; not all patients have all of them, and there may be central nervous system dysfunction (cerebellar vermis hypoplasia, agenesis of the septum pellucidum), renal anomalies (horseshoe kidney, hydronephrosis), variable heart disease, including hypertrophic obstructive cardiomyopathy, and distal limb defects such as syndactyly [41]. Most patients die in the late prenatal period or in the first months, from respiratory or heart failure; the oldest survivor described was 11 years old, with severe developmental delay, gastrostomy and nocturnal ventilation [41]. Following identification of an RBM10 variant, renal ultrasound, brain MRI and spinal ultrasound and serial echocardiographic surveillance for hypertrophic cardiomyopathy have been suggested [41]. The 2026 review counted 9 subjects with RS and an RBM10 variant [7]. The same review found, with 1 to 3 subjects per gene, variants in ATN1 (expansion; CHEDDA), ATP2B1, B3GALT6, BMP2, BMP4, BMPR1B, CREBBP, DPF2 (Coffin-Siris), EFTUD2, EIF4A3, FOXC2 (lymphoedema-distichiasis), GRHL3 and IRF6 (Van der Woude), HIST1H1E (Rahman), KIF15 (Braddock-Carey), KMT2D (Kabuki), LMNA, LOXL3, MED12, MED13L, MYMK (Carey-Fineman-Ziter), PGM1 (congenital disorder of glycosylation), SLC26A2, TCOF1 and ZC4H2 (Wieacker-Wolff in females) [7]. In the Minneapolis series, in addition to Stickler syndrome, central hypoventilation, Duane syndrome, Cornelia de Lange syndrome, Emanuel syndrome, Gordon syndrome, Moebius sequence and Nager syndrome were diagnosed [73].

6.7. Discriminating elements in practice

The deformation sequence due to oligohydramnios is distinguished by the history and course, with normalisation of mandibular size by about 2 years [2]. The mandibulofacial dysostoses are distinguished by the ear and eyelid findings, and examination of the parents is necessary, because Treacher Collins syndrome has highly variable expressivity and may be almost non-penetrant [2]. The oculo-auriculo-vertebral spectrum is distinguished by asymmetry and mainly unilateral involvement of the ear, oral structures and mandible [2]. In skeletal dysplasias, radiographs of the spine, scapulae, ribs and long bones are decisive, and in campomelic dysplasia a karyotype is mandatory in all children with a female phenotype [4, 34]. The combination of RS with myopia points to Stickler syndrome, and the Cleft Collective considers ophthalmological and genetic screening for Stickler syndrome mandatory in all RS [22, 77].

7. Treatments and Risk Management Strategies

7.1. Principles and organisation of care

Treatment of RS is supportive and directed at the compromised functions: breathing, feeding and growth, palate and speech, hearing, dentition and facial growth, and the psychological well-being of the child and family. There is no targeted drug. The 2016 consensus establishes that management of all newborns with RS begins with assessment and treatment of the obstruction [4]. Smith's textbook stresses that there is no consensus on the best approach and orders the options by increasing invasiveness: prone positioning, nasopharyngeal airway (NPA), nasoesophageal intubation, tongue-lip adhesion (TLA), MDO and tracheostomy [1]. The 2024 European guideline also includes the pre-epiglottic baton plate (PEBP, also called the Tübingen palatal plate, TPP) and continuous positive airway pressure (CPAP), and recommends that decisions be taken by a multidisciplinary team after continuous oximetry, clinical assessment, PSG and FNE [3].

The European guideline proposes national reference centres (one centre per 10 million inhabitants), with a care coordinator (often a nurse specialist), a round-the-clock service, a transition protocol to adulthood, systematic outcome assessment and collaboration with patient organisations [3]. The core team includes a perinatologist or neonatologist, coordinator, fetal medicine specialist, paediatric intensivist, dietitian or feeding specialist, paediatric anaesthetist, otorhinolaryngologist and audiologist, speech and language therapist, nurse specialist, cleft or craniofacial surgeon, psychologist, respiratory specialist and orthodontist; a paediatric surgeon, cardiologist, neurologist, ophthalmologist, ventilation team, clinical geneticist, social worker and prosthodontist should be available; the mandatory facilities are a paediatric intensive care unit, a sleep laboratory and audiological assessment [3]. Since management is mainly non-surgical, overall responsibility often lies with the non-surgical specialties (paediatrics and intensive care, anaesthesia, neonatology) [3]. The Brazilian HRAC-USP protocol, in Bauru, applied to all RS regardless of severity, is based on nasopharyngoscopy, nasopharyngeal intubation and feeding facilitation techniques, often without surgery [10, 101].

In France, care is organised in the SPRATON rare disease reference centre, with a coordinating site at Hôpital Necker-Enfants malades (Paris) and 21 competence centres accredited in 2017, integrated into the TETECOU rare disease health network (malformations of the head, neck and teeth) [18]. The medical team includes paediatricians or neonatologists, paediatric pulmonologists specialising in sleep, otorhinolaryngologists, geneticists, anaesthetists and paediatric plastic or maxillofacial surgeons, in collaboration with speech and language therapists, orthodontists, paediatric dentists, psychomotor therapists and psychologists [18]. A newborn without a prenatal diagnosis is in most cases transferred to the neonatal unit of the regional university hospital or to a competence centre; without signs of poor functional tolerance, the infant may be referred directly to the maxillofacial surgeon of the competence centre, who directs the infant to the paediatrician of the reference group [18].

7.2. Delivery and neonatal stabilisation

When there is prenatal suspicion, delivery should take place with a doctor experienced in the difficult neonatal airway present; in high-risk cases (severe micrognathia with an inferior facial angle below 49°, glossoptosis with a tongue shape index above 80%, polyhydramnios, CP or suspected syndromic RS), the guideline considers daytime delivery or a planned caesarean section with an obstetrician, anaesthetist, neonatologist, otorhinolaryngologist and neonatal intensive care unit [3]. The possible immediate interventions are NPA, laryngeal mask and orotracheal intubation, with videolaryngoscopy [3]. An infant in severe acute distress is stabilised according to standardised paediatric resuscitation algorithms, in consultation with a reference centre [3]. The French PNDS recommends that delivery of a fetus with suspected RS take place in a level 3 maternity unit, with an otorhinolaryngologist or paediatrician able to intubate in the delivery room and, ideally, integrated into the SPRATON national network [18]. The clinical case in section 10.2 shows that intubation may fail with direct laryngoscopy, videolaryngoscopy, laryngeal mask and lighted stylet, and that associated tongue-base lesions may make the obstruction multilevel [95].

7.3. Non-surgical treatment of the obstruction

The algorithm of the European guideline (appendix 6.1) starts from clinical observation, oximetry and a baseline PSG, which may not be possible in severe or life-threatening obstruction, and stratifies by severity: in mild obstruction (OAHI below 5) observation is maintained; in moderate obstruction (OAHI 5 to 10) prone or lateral positioning is used with reassessment; in severe obstruction (OAHI above 10), or if positioning fails (OAHI above 5), formal airway assessment by a multidisciplinary team and non-surgical respiratory support follow (NPA, CPAP, high-flow nasal cannula or PEBP); in life-threatening obstruction, the infant is intubated and the same assessment is carried out [3]. Surgical options (MDO or tracheostomy) are reserved for failure of non-surgical support (OAHI above 5), prolonged dependence on that support (6 to 12 months) or dependence on intubation for more than 2 weeks; TLA may be an intermediate option in experienced centres; the OAHI threshold of 5 is for guidance and PSG scoring should be weighed together with clinical data [3]. The European guideline recommends including non-surgical interventions (prone positioning, NPA, PEBP and/or CPAP) in the treatment algorithm, starting with them in mild to moderate OSA, training parents and carers in handling the NPA and nasogastric tube (NGT), considering endotracheal intubation as a rescue and bridging measure, escalating treatment if OSA is not controlled and reassessing as a team if non-surgical treatment is still needed at 6 to 9 months [3]. In the guideline's Delphi panel, there was consensus that the supine position may cause sleep-disordered breathing (90%), that the risk of sudden infant death syndrome (SIDS) should be weighed against the benefit of prone positioning (100%), that prone or lateral positioning with continuous oximetry, including at home, is an adequate first line (85%) and that, in the absence of demonstrated superiority, the choice should be based on efficacy and local experience (75%) [3]. In the first round, the statement that prone positioning may increase the risk of sudden death had only 45% agreement; the panel discussed that the issue is not the existence of risk, but the balance between the risk and the respiratory benefit, and the reformulation in those terms achieved consensus in the second round (February 2023) [3].

Positioning. Lateral or prone positioning is the most widely used first line and is most effective in mild obstruction; it loses efficacy after 4–6 months, when the infant changes position [3]. In 12 studies with 766 patients and variable definitions of success, positioning sufficed in 27.9%–72.2% of isolated RS, 41.7%–100% of non-isolated RS and 47.6%–100% of mixed cohorts [3]. PSG data show that positioning reduces but rarely eliminates OSA: in Helsinki, positioning alone was considered sufficient in 77.6% of 67 patients, but more than 75% had persistent moderate to severe OSA (OAHI above 5) [3]. The Tübingen randomised crossover trial (29 infants, 21 assessed in both positions) found no significant difference in the obstructive apnoea index between supine and prone positions in the main analysis; 13 of 21 benefited from the prone position and 8 worsened, with no interaction with syndromic status [102]. The guideline recommends trying the lateral position first and then the prone position, with oximetry [3].

The risk of SIDS is the point of greatest divergence. In healthy infants, the prone position is associated with a risk of SIDS about 13 times greater; the Tübingen group uses it only transiently after admission and does not discharge infants with a recommendation to sleep prone [93]. A systematic review of 23 studies (875 patients treated with positioning) recorded 6 deaths during positioning, none directly attributed to the position, with lower mortality in recent studies [103]. In France, the national SIDS registry of the last 10 years identified 1 case in a child with RS, who died awake in the supine position, and the survey of the reference centres (about 3000 children followed over 30 years) identified 2 cases; the authors concluded that the prone position may be offered to selected patients with strict precautions, an observational conclusion in which the absence of excess cases does not demonstrate safety [104]. About half of the centres worldwide use the prone position as first line [104].

The 2021 French PNDS makes prone positioning combined with EN the first line for all French teams: gravity brings the mandible and tongue forward, widens the pharyngeal space and reduces the obstruction, with success rates of 41% to 69% in the studies cited; it cautions that, during sleep, the position reduces the visible signs of respiratory difficulty and is associated with a greater risk of sudden death, so it requires maximum safety conditions, and that, once these conditions are met, a home monitor is not indicated, although some teams prescribe one; a non-nutritive dummy is a useful complement in minor obstructions [18]. The PNDS refers to the prone position "with or without slight head-up tilt", but the precautions sheet of the Paris reference centre, revised in June 2025, excludes the head-up tilt and lists the conditions required for discharge in the prone position, after explanation and parental agreement: a firm, wide mattress (60 × 120 cm cot), no Moses basket; a well-stretched bottom sheet; no cot bumper; head without support; no soft toy, pillow, towel or item of the mother's clothing; child without blanket or duvet, with pyjamas suited to the temperature or a correctly sized sleeping bag, and in summer pyjamas only; room temperature up to 20 °C; no tobacco, alcohol or drugs in the room; and never sleeping with the baby prone in the parents' bed. The same sheet adds that, according to the 2024 European guideline, an oximeter may be offered during sleep [3, 105]. There are therefore three positions: the European guideline prefers trying the lateral position first and weighing the risk of SIDS; the Tübingen group does not discharge infants in the prone position; and French practice maintains home prone positioning with strict rules [3, 18, 93].

Nasopharyngeal airway. The NPA is a small-calibre tube inserted through the nose until it passes the tongue base; if it is too short it does not bypass the obstruction, if too long it stimulates the epiglottis and causes vomiting or coughing; the position is correct when air comes out through the tube on expiration and little saliva or milk passes during feeding; it easily becomes blocked with secretions and parental training is needed [3]. In 10 studies (538 patients), success was 60%–100% in isolated RS and 50%–100% in mixed cohorts [3]. At Great Ormond Street, the NPA was effective in 63 of 77 patients with moderate to severe obstruction on PSG, removed after a median of 8 months with PSG every 2 months; 14 patients with persistent severe obstruction underwent tracheostomy, and 9 of these were decannulated without further treatment at a mean age of 3 years [3]. In Bauru, 107 severe RS treated with hospital and home nasopharyngeal intubation kept it for a mean of 57 ± 38 days, without complications or other measures; reassessment took place every 15 days up to a maximum of 6 months [3]. The NPA allows discharge with home support [3]. The French PNDS brings together other series: the first success in 12 patients (London, 1982); 48 patients over 20 years at Sainte-Justine hospital in Montreal; in Minneapolis, 74 RS over 20 years, of which 38 needed relief of the obstruction, 14 with NPA, 18 with MDO and 6 with tracheostomy; in Ireland, 69 RS, 27 requiring intervention, 16 treated successfully with NPA, 2 failures that needed glossopexy and NIV and 8 tracheostomies; in the Brazilian series that includes the 107 Bauru patients cited above (223 RS, 73% isolated), mean admission of 18 days, tracheostomy avoided or decannulation in all, gastrostomy in 15% and zero mortality; and, in France (2018), 15 newborns with satisfactory ventilation and saturation after a mean of 64 days of NPA, with improved capnia and weight gain at the cost of attentive care and parental participation [18]. These series have no control group and the criteria for success are sometimes not very rigorous; the advantages of the NPA are minimal morbidity, low cost, contact between parents and child and early discharge once the patient has been stabilised and the parents trained [18]. For infants with excessive tone who tolerate the ventilation mask poorly or who are obstructed when awake, the PNDS considers the NPA an alternative to NIV [18].

Continuous positive pressure. At Necker, among 44 patients, 9 received CPAP (7 from birth for severe obstruction and 2 at 7 and 10 days for moderate obstruction on PSG), with criteria of OAHI above 10 per hour, desaturation index above 15 per hour, minimum SpO2 below 90% or transcutaneous CO2 above 50 mmHg; the median duration in isolated RS was 1.6 months and overall success was 9 of 13 [3]. In another study, CPAP resolved the obstruction in 3 of 11 (27.3%), and the rest were switched to NPA [3]. The disadvantages are the difficulty of achieving a seal and the constant pressure on the immature craniofacial skeleton, with evidence of midface hypoplasia associated with the mask; a high-flow nasal cannula is an alternative, without equivalent evidence [3, 93]. The combination of NGT and CPAP requires attention to pressure injuries, air leak and gastric distension [3].

The French PNDS gives more weight to non-invasive ventilation (NIV) by nasal mask with continuous positive pressure, which the Necker team uses as first line in obstructive RS [18]. In the initial work of Fauroux and colleagues, 7 infants meeting the criteria for relief of the obstruction (6 with an associated malformation), but able to breathe autonomously for a few minutes, had work of breathing on average six times greater than normal, which helps to understand the growth deficit; NIV significantly increased mean saturation, was used for a mean of 8 hours per day, required 1 to 2 weeks of adaptation, all were discharged on NIV, none was tracheostomised and the mean duration was 16 ± 12 months (2 children were still using it at 13 and 39 months) [18]. Amaddeo and colleagues (2016), in 40 newborns, successfully treated with NIV 5 of the 9 with severe obstruction and the 4 with moderate obstruction [18]. The main problem is tolerance of the mask by irritable infants and by the parents, which requires patience and trained nurses; the PNDS observed no adverse effects, in particular maxillary deformation, which it attributes to the short duration (a few months) [18]. This statement differs from the European guideline, which refers to evidence of midface hypoplasia associated with the mask and points to pressure on the immature craniofacial skeleton as a disadvantage of CPAP [3].

Endotracheal intubation. It is a rescue measure and a bridge to another treatment; in small studies, extubation was successful in 12 of 17 isolated RS (70.6%) and in 17 of 28 infants from a mixed cohort after failure of positioning or NPA (60.7%), with a risk of subglottic stenosis with prolonged intubation [3].

Pre-epiglottic baton plate. The PEBP is a palatal plate with an extension (velar spur) that extends behind the tongue base and displaces it forward, combined with feeding training; it is adjusted about every 3 months, because the maxilla grows [3]. In a randomised crossover trial of 11 isolated RS with a mean age of 3 days, the MOAI fell from 13.8 to 3.9 per hour after 48 hours with the PEBP and remained at 14.8 with the plate without extension [3, 93]. In a two-centre study of 49 patients, the median MOAI fell from 15.9 to 2.3 per hour at discharge and to 0.7 at 3 months, without the need for mechanical ventilation or tracheostomy [3]. In 303 isolated RS from Tübingen (1998–2019), with indication by MOAI of 3 or more, success was 99.0%; the MOAI fell from 9.0 to 0.9 per hour at 3 months, NGT use went from 55.1% to 13.7% and the weight z score, which was −1.17 at discharge, reached −0.44 at one year, with exclusive oral feeding [3]. In 6 studies (814 patients), success ranged from 78.0% to 100% [3]. The guideline notes bias due to application to patients who would have been stable with positioning, availability limited to centres with orthodontic experience and risks of incorrect sizing, pressure injuries and device failure [3]. Outside Germany, uptake is limited: in a North American survey (122 responses), only 9.0% of centres offered the PEBP, and the barriers were doubts about efficacy (54.9%), lack of staff to manufacture it (33.6%), family acceptance (27.9%), potential for harm (21.3%) and billing (15.6%) [106]. In Seattle, an orthodontic airway plate pathway started in 2023 reduced the mean OAHI from 71 to 7.9 per hour in 9 patients, all weaned from respiratory support [107]. The Necker hospital published in 2026 the protocol for implementation outside the centre of origin and the first series [108]. A systematic review of 10 studies (598 patients, 483 isolated) concluded that the plate improved the obstruction and avoided tracheostomy in most, with a small proportion of syndromic patients needing surgery [109]. The French PNDS describes the Tübingen plate as a palatal part covering the hard palate, the cleft and the alveolar ridges, with a velar extension of about 3 cm whose tip reaches down to the vallecula, worn 24 hours a day and removed once a day for cleaning, held in place by adhesion, suction and adhesive cream, with extraoral bows fixed to the face with adhesive tape and generally replaced after 3 months, if a lesion of the alveolar ridges appears or if the sleep study worsens [18]. It also mentions the three types of Würzburg plate (wire extension, plastic extension and hollow pharyngeal tube), the retrospective series of Buchenau and colleagues (207 RS, 59% syndromic) with reduction in mixed and obstructive apnoeas and improved feeding, and the results presented by Tübingen in 2019: superiority over a sham procedure in a randomised trial, intellectual development within the reference range, MOAI close to normal in 83% and, among 443 infants treated (129 syndromic), final tracheostomy in 23 (5%), all with syndromic RS [18]. The French authors consider these data very promising, but difficult to export, because of the handmade nature of the plates and dependence on a team that handles them regularly [18].

The French PNDS stresses that a newborn with isolated RS, or with an associated bone disease, has a normal intellectual prognosis, so hypoxia and hypercapnia must not compromise that prognosis and the method of relieving the obstruction must be fully effective [18]. It recalls that there is no international consensus on algorithms, that a North American team published a mainly surgical algorithm in 2018 and that Europe favours non-invasive strategies: NPA in England, palatal plate with pharyngeal extension in Germany, NIV and NPA in France [18]. The order of treatments of the French reference centre is: prone positioning as first line; nutritional support and/or suspension of bottle feeds to get through a phase of worsening during feeds; NPA or NIV, depending on local experience or after failure of the first technique; TLA, only in the few teams with experience and in children whose obstruction depends mainly on glossoptosis; and tracheostomy after failure of the previous strategies [18]. The Necker respiratory algorithm is summarised in the following table.

Source of the table: algorithm of the Necker reference centre, reproduced in the PNDS [18]. The algorithm uses SpO2 and PtcCO2 criteria and an AHI above 10 per hour, whereas the European guideline stratifies by OAHI with a threshold of 5 per hour; both agree on progressive escalation and on reserving surgery for failure of non-surgical treatment [3, 18].

7.4. Surgical treatment of the obstruction

The European guideline recommends considering MDO or tracheostomy for OSA that does not respond to non-surgical treatment, after excluding obstructions below the tongue base; it accepts TLA as an alternative in experienced centres, but advises that breathing and feeding outcomes are better with MDO, in a shared decision; it recommends three-dimensional virtual planning when an internal device is used; it recommends preferring tracheostomy to MDO in syndromic patients, those with low birth weight and those with neurological compromise; and it recommends overcorrection in MDO, because of the intrinsic disturbance of mandibular growth and the effect of the surgery itself [3]. In the event of persistence or recurrence, the decision is reviewed by the team, and MDO may be used to end prolonged non-surgical treatment, such as CPAP, or to allow decannulation [3]. The French PNDS recalls that the main candidates for the more invasive interventions are patients with syndromic RS and that, in a recent study, only 10% of isolated RS needed a surgical approach [18].

Mandibular distraction osteogenesis. MDO lengthens the mandible with an internal or external device and has been used since the 1990s [3]. The systematic review of 49 studies gives a success rate of 94%, and the patients who needed tracheostomy after MDO were non-isolated with an AHI above 15 [3]. In a series of 177 patients, the AHI fell from 43.3 to 7.0 with MDO, and the main predictor of surgery was an AHI above 20; tracheostomy after MDO was needed in 8.8%, associated with low birth weight and neurological compromise [3]. In a review of 408 patients, primary success was 97.6% in isolated RS and 90.7% in syndromic RS, and failures were not associated with age; an American paediatric surgical database (208 patients) recorded 30-day adverse events in 7.3%, more frequent before 1 year of age (26.1% versus 10.8%) [3]. The complications described are pin-site infection (6.3% in a meta-analysis of 711 cases), open bite (3.0%), nerve injury (2.5%), hypertrophic scarring (2.3%), permanent injury to the first molar (48% in one series) and weakness of the depressor labii inferioris [3]. A Pittsburgh series (55 infants, MDO before 1 year, mean follow-up of 6 years) found palsy of the marginal mandibular branch of the facial nerve in 38.2%, permanent in 29.1%, well above the 1%–15% usually reported [110]. For feeding, a review of 300 patients showed exclusive oral feeding at 12 months in 82%, 93.7% in isolated and 72.9% in syndromic RS [3].

Tongue-lip adhesion. TLA can be performed in the first weeks of life; it is usually taken down at 6–8 months or at palatoplasty [3]. In a review of 268 patients, success was 81.3% (91.5% in isolated and 79.8% in syndromic RS), with complications in 13.8% (dehiscence in 8.2%) and additional surgery in 12.3% [3]. There are no data on the effect of TLA on long-term mandibular growth [3]. The French PNDS describes the technique (anchoring of the tongue to the mucosa and muscle of the lower lip, in the first weeks of life, without correcting the micrognathia) and the complications: spontaneous dehiscence, the most frequent (17%–28%), tongue lacerations, injury to Wharton's ducts, infections and scar deformities, speech articulation disorders, tongue coordination problems, reoperation, aspiration pneumonia and dental malposition [18]. In the systematic review by Viezel-Mathieu and colleagues (2018; 268 patients, 35.1% isolated, 35.1% syndromic and 29.8% associated), the complication rate was 13.8% with TLA, versus 23.8% with MDO and 37.5% with tracheostomy, and 5 patients later needed tracheostomy and 5 MDO [18]. The transient language disorders described by Leblanc resulted from a different technique, with fixation of the tongue to the mandible and ankyloglossia [18]. For most French teams, TLA no longer has a role, although Lille, Marseille and Tours still perform it; feeding comfort after TLA has been little studied [18]. Subperiosteal release of the floor of the mouth (Delorme procedure), developed at Sainte-Justine hospital and used only in Quebec, is not recommended by the PNDS [18].

The French PNDS describes MDO in three phases: latency of 0 to 2 days after the osteotomy; activation, generally at 1 mm per day (0.5 mm in the morning and 0.5 mm in the evening); and consolidation, for 6 to 12 weeks, for mineralisation of the new bone [18]. It cites a Boston series (19 RS treated before 3 months and followed for 3 years), a Canadian series with positive effects on feeding and reflux in 11 children and a Dutch series of 12 children treated with resorbable internal distractors (surgery at a mean of day 32, mean interalveolar gain of 18 mm, extubation 8 days and discharge 17 days later), with fewer complications, although there are almost no long-term follow-up studies and MDO may not relieve multilevel obstruction [18]. External distractors give better control of movement in two planes, and internal ones have fewer infectious and aesthetic drawbacks, but are more difficult to place and to activate; the list of complications includes temporomandibular ankylosis, fracture, injury to tooth germs, malunion, failure of traction, facial paresis, abscesses and asymmetric mouth opening [18]. The PNDS does not recommend early MDO in France, except in RS with major mandibulofacial dysostosis in which retrognathia has been shown to be solely responsible for the obstruction and in situations in which mandibulofacial growth is known to be insufficient [18].

Comparison between MDO and TLA. The guideline's synthesis gives resolution of OSA in 96% with MDO and 45% with TLA, secondary treatment in 2%–6% and 22%–45%, tracheostomy avoided in 97% and 89%, respiratory failure after TLA in 21%–29%, dehiscence after TLA in 30%–55% and MDO complications in 6.7%–48%, with low quality of evidence [3]. Neonatal MDO (before 28 days) did not have lower success [3]. In a Boston comparison, resolution of OSA occurred in 22 of 24 patients with MDO (92%) and in 9 of 19 with TLA (47%) [3]. Quality of life improves after both, with no difference [3]. For feeding, according to the data summarised by the PNDS, weight gain and long-term growth did not differ between MDO and TLA, but, in two retrospective comparative studies, more patients reached full oral feeding after MDO, with a faster return to oral feeding, discontinuation of supplements and fewer gastrostomies [18].

Tracheostomy. It is the most effective intervention for severe OSA, but prolonged cannulation interferes with speech and language and is associated with granulations, suprastomal and tracheal collapse, subglottic stenosis, haemorrhage, fistulae and bronchopulmonary infections [3]. In the series of 177 patients, 80% of those with a tracheostomy needed gastrostomy or surgical feeding access and exclusive oral feeding was reached at a mean of 763 days; 19 of 33 patients with a tracheostomy who later underwent MDO were decannulated [3]. The French PNDS considers tracheostomy the gold standard for treating major upper airway obstructions, easy to institute and radically effective, and explains that it has so far been the technique most used in France in severe RS, owing to the quality of social protection and the culture of home care and therapeutic education; the problem is not efficacy but the drawbacks (morbidity, burden of home care, risk of mortality and, in some countries, cost), and its long-term efficacy was shown in a prospective study of psychomotor, cognitive and language development and feeding disorders in 39 severe RS (isolated and with Stickler syndrome) [18]. There are no studies on the complications of tracheostomy specifically in infants with RS [18]. The Brazilian textbook indicates tracheostomy in all cases of severe obstruction on nasopharyngoscopy, and the GeneReviews chapter on Stickler syndrome notes that, in most, the micrognathia becomes less prominent with time and allows the tracheostomy to be removed [10, 12].

Recent comparative evidence. A meta-analysis of 13 studies compared MDO and tracheostomy: MDO was associated with a better airway outcome (OR 10.72; 95% CI 1.97–58.44), less gastrostomy (OR 0.09; 95% CI 0.04–0.18) and lower cost, with confounding by syndromic status [111]. A two-centre cohort from Boston and Tübingen (114 infants treated before 6 months) compared MDO (31) and PEBP (83): the reduction in OAHI was 98% and 94%; PEBP was associated with fewer surgical feeding tubes and faster early growth; in moderate obstruction the respiratory outcomes were similar, and in severe obstruction MDO more often achieved resolution (level III study) [112]. A systematic review of 23 published algorithms found an AHI above 20 to be a frequent threshold for surgical escalation and a preference for MDO over TLA in 16 [113]. A Colombian algorithm proposal, based on a review using the GRADE system (Grading of Recommendations Assessment, Development and Evaluation) of 23 articles published between 1985 and 2016, is another example of these protocols [114]. A 2019 international survey showed that US surgeons operate earlier and prefer MDO, whereas outside the USA the various techniques are used to a similar extent [115]. In Scotland, the NPA is the first line and only 7% of patients undergo surgery; in a multistate US study, 77% were treated surgically [17].

7.5. Feeding and nutrition

Almost all children with RS will have a combination of feeding problems, and insufficient growth is a marker of inadequate airway treatment [3]. The European guideline recommends standardised feeding assessment protocols at presentation, with frequent monitoring early in life; prenatal counselling on breastfeeding; standardised growth charts; assessment by a dietitian and by a speech and swallowing therapist in all patients, isolated and non-isolated; recording of intake, weight, length and head circumference; and closer monitoring during surgical interventions [3]. It values bonding and breastfeeding, with an early lactation consultant, bottles and teats with variable flow, and a low threshold for treating reflux and swallowing dysfunction on clinical grounds; pH monitoring and motility studies are reserved for complex refractory cases [3]. The Oxford Desk Reference and the Bauru protocol describe the NGT as an initial measure, which in some infants is enough to keep the tongue forward and the airway open for days or weeks, and facilitation techniques (positioning, modified teats) [3, 46].

The text of the guideline chapter recommends using an NGT when oral intake is unsafe or insufficient for growth and considering high-calorie formulas if growth remains insufficient with standard feeding; the summary of the same guideline is less restrictive and proposes high-calorie formulas as first line and an NGT when intake remains insufficient with modified formulas [3]. This internal divergence should be known; the practice that follows from the more detailed chapter text is to reserve high-calorie formula for insufficient growth. In Bauru, 23 isolated RS received formula supplemented with 5%–7% glucose polymers and 3%–5% medium-chain triglycerides (9 patients) or standard formula (14), and the supplemented group had earlier clinical improvement, shorter nasopharyngeal intubation and greater weight gain; this is a study finding, with low quality of evidence, and formulas should be prescribed by a dietitian or paediatrician, with monitoring for osmotic diarrhoea and electrolyte and trace element imbalances [3]. The NGT improved weight gain in 53%–100% of isolated RS and 0%–83% of non-isolated RS, in four studies with high risk of bias; in one series, 10 of 19 isolated RS with an NGT progressed without gastrostomy, whereas all 10 non-isolated RS needed gastrostomy [3]. Gastrostomy is indicated when support is expected to be permanent or prolonged and in children with UAO who fail respiratory interventions; there are no studies on its effects [3]. The NGT, NPA and MDO may cause oral aversion, and MDO and TLA are associated with temporary perioperative weight loss, with faster recovery after MDO [3]. With the PEBP and feeding training, about 85% of infants have the NGT removed during the initial admission [3].

The French PNDS starts from a physiological principle: sucking-swallowing-breathing is a reflex integrated in the brainstem and, by definition, cannot be retrained; time spent feeding has no training effect, and sucking is most effective in the first 10–15 minutes, so weak sucking should be compensated for rather than forced [18]. The facilitation methods proposed are to wait until the infant is hungry and respond to its demand, keep it awake during the bottle feed (avoiding it being too snugly held in the mother's arms), use long, soft rubber teats with wide slits or a Haberman teat with a valve, thicken the milk (carob flour) so that flow through the slit teat is slow, and keep the head in line with the body, without posterior extension of the neck, to facilitate laryngeal closure; evidence for their isolated efficacy is scarce, as a Cochrane meta-analysis shows [18]. In the maternity unit, the document recommends relieving the guilt of the mother for whom breastfeeding will not be possible, using soft rubber teats and not wide-based ("plateau") teats, giving expressed breast milk, if the mother wishes, thickened in the bottle (in France, with Gumilk®, a carob thickener) or, alternatively, a milk thickened with carob starch, feeding the infant in a well upright position, putting it to sleep in the prone position under safe conditions and monitoring SaO2, cardiorespiratory rhythm and obstruction when crying, at rest in the supine, lateral and prone positions, awake, asleep and during feeds [18].

Oral feeding should be temporarily suspended, for a few days or weeks, in two rare situations: when feeding causes aspiration and when it worsens respiratory difficulty to the point of risking invasive relief of the obstruction (NIV, tracheostomy); in all other cases, feeding is allowed without forcing, because it satisfies the child and the mother, and infants who fed partially in the first weeks had shorter EN than those who did not [18]. For infants with moderate weight gain, the PNDS text advises enriching the milk formula with 3% to 5% glucose polymers and 1% to 3% lipids (medium-chain triglycerides and essential fatty acids), whereas the summary of the same document indicates 5% glucose polymers and 1% lipids [18]. If enrichment and facilitation are not enough for normal weight gain, EN by NGT is required, exclusive if there is aspiration or, preferably, partial, as a complement to bottle feeding; it is usually transient and well tolerated and can be carried out at home once the parents have been trained [18]. To prevent secondary behavioural disorders, the PNDS favours bolus EN, avoids stress and lets the child explore the face and mouth and taste the foods it wants, with psychomotor, speech and language therapy or psychotherapeutic support as appropriate [18].

The PNDS proposes gastrostomy in only two situations: when antireflux surgery is required because of failure of medical treatment of reflux, in which case the gastrostomy is placed during Nissen fundoplication; and when the need for EN extends beyond 9 months with poor tolerance of the tube [18]. In the reference centre's experience, most of these children had syndromic or associated RS; the isolated RS cases that needed Nissen with gastrostomy were all grade 3, had a tracheostomy and uncontrolled vomiting; percutaneous gastrostomy without antireflux surgery is discussed in prolonged EN with tube intolerance in a child with neurological disease and without reflux, and is not useful in isolated RS [18]. Regurgitation, reflux and vomiting are very frequent but, according to the PNDS, do not correspond to classical gastro-oesophageal reflux due to inappropriate relaxations of the lower oesophageal sphincter: they result from disorders of oro-oesophageal motility and mechanical changes in intrathoracic and abdominal pressures caused by the obstruction [18]. For this reason, pH monitoring and upper gastrointestinal contrast studies are often normal and of little use; oesophageal manometry shows changes in lower sphincter tone and wave propagation, but, in the reference centre, did not predict the duration of EN and is not indicated; the contrast study is only requested before antireflux surgery [18]. Medical treatment is the usual one (thickening with starch and a proton pump inhibitor, PPI), with EN slow enough not to worsen vomiting, and fundoplication is reserved for reflux that does not respond and worsens respiratory status [18]. The PNDS does not give doses; the reference PPI doses are in section 7.10.

7.6. Palatoplasty, speech and hearing

The timing of palatoplasty in RS is later than in isolated CP, because of the airway: the Brazilian textbook indicates palatoplasty from 12 months and, in children with a tracheostomy, before decannulation, whereas for clefts in general the Oxford Desk Reference gives repair of the posterior palate at 6–8 months [10, 46]. Published studies show mean ages of 12 to 18 months in RS and a mean delay of 2 to 5 months compared with isolated CP [3]. Palatoplasty may precipitate respiratory compromise due to oedema of the tongue and palate: in Utrecht, 14 of 53 RS versus 9 of 74 isolated CP had postoperative respiratory difficulty, with an OR of 4.16 (95% CI 1.17–15.99) in non-isolated RS, without clear benefit from deferring until after 12 months and without predictive value of preoperative screening with a palatal plate [116]. In other series, respiratory complications were 6.8% (RS) and 7.7% (isolated CP) when surgery was performed only with an AHI of 5 or less without CO2 retention, 30% with rapid resolution when a normal PSG was required, and 7% in Tübingen after PEBP, with temporary prophylactic NPA in 11.2% [3]. The guideline concludes that palatoplasty is safe in terms of breathing with an AHI below 5 without CO2 retention and/or a normal PSG with or without a plate; in the Delphi panel, there was consensus for PSG before palatoplasty (70.5%), for using PSG in deciding on the type and timing of palatoplasty (70.5%) and speech surgery in children with sleep-disordered breathing (82%), and for repeating PSG at least 3 months after palatoplasty or speech surgery (82%) [3]. The initial recommendation to operate only with a normal PSG was not approved by all parties, and in the first round the statements on PSG before and after palatoplasty had only 55%–65% agreement, because some members considered clinical assessment sufficient [3]. The guideline's algorithm proposes, in a child with CP, a PSG with simulated cleft closure (palatal plate): if the OAHI is above 5, palatoplasty is deferred and PSG repeated 6 to 8 weeks later; if simulated closure is not possible, deferring palatoplasty is considered with an OAHI above 5 [3].

The French PNDS acknowledges that there is no consensus on the cleft closure strategy in RS and that comparative studies are rare [18]. The aim is functional closure that prevents nasal regurgitation of food and allows the best possible phonation, without rhinolalia and without compromising maxillary growth [18]. The technique recommended by the reference centre, and the one most practised in France, is Sommerlad intravelar veloplasty, at about 6–7 months (5–7 months in the 2018 series of Caillot and colleagues), with closure of the bony palate at the same time if the cleft is narrow or deferred to 15–18 months (12–16 months in the same series) if it is wide, without lateral incisions [18]. In partial clefts, repair is done in one stage, because the very vertical palatal shelves give the mucoperiosteum enough length once it is horizontalised; in complete and wide clefts it is done in two stages; in very wide clefts lateral incisions and a vomer flap may be added [18]. Uranostaphylorrhaphy (midline suture of the palate and velum at about 9 months) is no longer recommended, because of the risk of maxillary hypoplasia linked to elevation of the mucoperiosteal flaps, and modified Furlow palatoplasty with intravelar veloplasty, practised in Lille, now has few proponents; in that series of 92 patients operated on over 4 years, 4 had postoperative respiratory complications [18]. The ages proposed by the PNDS are earlier than the means of 12 to 18 months in the studies reviewed by the European guideline and than the 12-month threshold of the Brazilian textbook [3, 10, 18].

The guideline recommends considering avoiding pharyngoplasty (pharyngeal flap or sphincteroplasty), because it may induce OSA or cause it to recur, and, if it is considered for velopharyngeal insufficiency, informing the parents of the alternatives (intravelar plasty, buccal flap to lengthen the soft palate) [3]. The French PNDS describes velar insufficiency as frequent and plans secondary surgery between 5 and 6 years, after several years of speech and language therapy, with several possible techniques: secondary intravelar veloplasty, superiorly based (Sanvenero-Rosselli) or inferiorly based velopharyngoplasty, Orticochea dynamic sphincteroplasty and autologous fat injection into the nasopharynx or velum [18]. Pedicled velopharyngoplasties may worsen an obstructive syndrome, so resolution of the respiratory disorders should be awaited and, at the slightest doubt, a preoperative sleep polygraphy requested; for this reason, some authors prefer dynamic sphincteroplasty or fat injection [18]. This preference contrasts with the European guideline, which includes sphincteroplasty among the pharyngoplasties to be avoided [3]. For speech, it recommends early stimulation advice before and in the first weeks after palatoplasty, at least one formal individual speech and language assessment between 18 months and 3 years by a speech and language therapist specialised in clefts and VPD, regular reviews from 5 years until the end of facial growth (at least every 5 years without concerns and every 6 to 12 months with concerns), specialised velopharyngeal investigation (perceptual assessment, lateral videofluoroscopy and nasendoscopy) when VPD is suspected, and liaison with the tracheostomy therapist where applicable [3]. The guideline's summary refers to formal assessment "at 2 and 3 years" and screening "at least every 5 years", which differs slightly from the text of the guideline chapter [3]. In the French model, the first contact with speech and language therapy occurs at about 18 months, when the velum and bony cleft have been closed, in the form of group parental guidance (at Necker, up to 6 children), to reassure and inform families, prevent disorders of phonation, speech and language and start games for controlling intraoral pressure; a second session follows between 24 and 30 months, with referral to local speech and language therapy if there are hearing disorders, jargon or communication disorders, and to the speech and language therapist, psychomotor therapist and psychologist trio if feeding or developmental problems persist; after 3 years, the child is seen annually, or more often around secondary surgeries, until adulthood, with an interval of 2 years after 7 years if there is no disorder [18]. The cleft team should not discharge the child in childhood, because velopharyngeal function changes with growth and adenoid involution, and hearing surveillance (otitis media and hearing loss) is part of follow-up [3, 23].

7.7. Dentition and facial growth

The European guideline recommends a dental appointment at 5–6 years, with an orthopantomogram to screen for hypodontia, and a "bi-annual" review until skeletal maturity; the term is ambiguous in the original (six-monthly or two-yearly) [3]. It recommends centralising dental screening and surgery in reference centres, with orthodontics in shared care, and informing families early that functional (respiratory and occlusal) and aesthetic problems may arise at skeletal maturity, treatable with mandibular and maxillary advancement [3]. Complete mandibular catch-up growth is not expected; orthodontic-surgical correction takes place after the end of growth, and, given the frequent agenesis, the aim is facial harmony and reducing the recurrence of OSA in adulthood, with prosthetic options and implants, and not necessarily a class I occlusion [3]. Orthodontics in CP usually starts in the early mixed dentition, at 7–8 years [46]. In Stickler syndrome, mandibular advancement may be needed for persistent micrognathia with malocclusion [12]. The French PNDS brings the timetable forward: first paediatric dentistry appointment ideally between 4 and 5 years and first orthodontic assessment generally at 6 years (at 4 years, at the request of the maxillofacial surgeon, in severe or syndromic and associated RS), with clinical examination, orthopantomogram and frontal and lateral cephalograms to determine the occlusion in the anteroposterior (Angle classes I, II and III), transverse and vertical dimensions and the skeletal discrepancy [18]. Orthopaedic maxillary expansion may be proposed to correct dentomaxillary disharmony, followed later by alignment when all the permanent teeth are present, and severe cases require specific orthodontics before mandibular or, more rarely, maxillary advancement; at Necker, the appointment is repeated every 12–18 months [18]. Maxillofacial surgery follows the child from the neonatal period: neonatal appointment and one at about 6 months to schedule the first operation (at about 4 months, in the summary of the same document), impression of the maxilla in theatre at each operation, decision on one- or two-stage closure, postoperative appointment and appointments at 2 and 3 years and, thereafter, at least every 2 years until adulthood, with orthopantomogram, frontal and lateral cephalograms and three-dimensional photography at 6, at 12 and after 16 years (CT at 6 years if there is a problem) [18].

7.8. Psychosocial support and quality of life

In a survey of parents, half received psychological or psychosocial counselling and 74% of those who did not would have liked to; parents of children with RS report clinical levels of distress in 34.7%, higher in non-isolated RS and in the groups treated with NPA, MDO or tracheostomy [3]. The guideline recommends parental access to clinical psychology and social work linked to the cleft team, screening of parents at 1, 6 and 12 months with the Distress Thermometer for Parents, screening of the child, with or without CP, at 2, 5, 8, 12, 15 and 22 years with validated instruments, attention to complex social problems and early information about patient organisations [3]. It also recommends that parents have a central role, that parental assessment of the child's quality of life be screened with validated instruments and that patient-reported outcome measures be considered [3]. In Tübingen, 17 children with isolated RS treated with PEBP had, at 7–11 years, quality of life similar to that of children with CP and higher than that of European reference cohorts [117].

The French PNDS gives guidance on disclosure of the diagnosis, prenatal or postnatal: a dedicated appointment, with the involvement of a psychologist during or immediately after the appointment, explanation of the main features of the disease, the possibilities for treatment and follow-up and the existence of the patient organisation and social support, and a further appointment or telephone contact a few weeks later to go over the information again, which is difficult to take in amid the shock of the disclosure [18]. It provides for psychological support in two periods: in the first months, to support the parents in facing the diagnosis and therapeutic decisions that may seem aggressive and that transform the parental project, and, during the first year, to stimulate the child and compensate for the inhibiting nature of care; and in adolescence, a period of vulnerability; between the two, surveillance is maintained [18]. Psychomotor therapy is not mandatory for all teams, but is indispensable when RS is not isolated and is associated with a developmental disorder [18].

Therapeutic education of the parents has a central place in the French model: understanding the disease and recognising the obstruction clinically, learning the methods for facilitating sucking and, if necessary, home EN by NGT, and training in home airway management techniques (prone positioning, NIV, NPA, tracheostomy) [18]. At the coordinating centre there are three programmes approved by the regional health agency: home tracheostomy management (15 to 20 sessions of about 1 hour over 4 weeks of admission, with educational diagnosis, a theory session, observation and supervised practice of suctioning, care and cannula change, safety skills for accidental obstruction and decannulation, and a final cannula change carried out by the parents with the hospital-at-home team); home EN management (5 to 7 sessions of about 1 hour over 8 days, with checking of tube position, connection of feeds and changing of the tube or gastrostomy button, assessed using a grid); and NIV training, through the regional therapeutic education programme for chronic respiratory failure in children treated with NIV [18]. Follow-up appointments are grouped, whenever possible, into a multidisciplinary clinic or on the same day [18].

7.9. Particular features of the syndromic forms

In Stickler syndrome, patients with RS need immediate airway assessment by an otorhinolaryngologist and paediatric intensive care, as well as ophthalmological surveillance and prevention of retinal detachment [12]. In type II collagenopathies, respiratory failure and OSA may require CPAP; central apnoea requires reassessment for unrecognised cervical instability, and cervical spine imaging before anaesthesia is recommended [27, 35]. In campomelic dysplasia, anaesthesia carries increased risk and requires precautions if there is a cervical anomaly, and 46,XY children with female or undervirilised genitalia have an indication for gonadectomy because of the risk of gonadoblastoma, with no defined age [34]. In cerebro-costo-mandibular syndrome, ventilation is difficult because of the rib gaps; the number of absent ribs guides respiratory planning and follow-up includes scoliosis [38, 60]. In Catel-Manzke syndrome, respiratory support, cleft surgery and cardiac surgery are the main interventions [99]. In 22q11.2 deletion, the European guideline recommends investigating cardiac, urogenital, skeletal, gastrointestinal and ophthalmological anomalies and providing cognitive and psychological support, in addition to managing hypocalcaemia and immunodeficiency [2, 3]. MDO is, in several series, less effective in the syndromic forms, and the guideline recommends preferring tracheostomy in syndromic patients with low birth weight and neurological compromise [3].

7.10. Drugs and regulatory status

There is no pharmacological treatment for RS and no medicinal product authorised by the Food and Drug Administration (FDA) or the European Medicines Agency (EMA) for this indication; the sources consulted do not describe pharmacological doses specific to RS [3]. High-calorie formulas are nutritional interventions and not medicinal products labelled for RS (section 7.5). Treatment of reflux and comorbidities follows general paediatric recommendations and has not been studied specifically in RS in the sources consulted [3]. The only pharmacological treatment mentioned by the French PNDS is the PPI, combined with thickening, for reflux in children with RS, without a dose [18]. The reference doses are, by extrapolation, those of the joint North American and European paediatric gastroenterology guideline (NASPGHAN/ESPGHAN, 2018) for paediatric gastro-oesophageal reflux disease, and the US label of esomeprazole serves for comparison [118, 119]. The guideline recommends a PPI as first line for erosive reflux oesophagitis, a 4- to 8-week course for typical symptoms, reassessment if there is no response after 4 to 8 weeks of optimal treatment and regular review of the need for prolonged acid suppression, and recommends against using PPIs or H2 antagonists for crying or visible regurgitation in otherwise healthy infants; it suggests thickening feeds for visible regurgitation and vomiting, with carob bean or xanthan gum thickeners in breast milk, and not cereals [118]. PPIs have been associated, in case-control studies, with more infections (necrotising enterocolitis, pneumonia, upper respiratory tract infections, sepsis, urinary tract infections and Clostridium difficile infections), which reinforces the recommendation of short courses [118].

Oesophagitis healing rates in paediatric trials were consistently high with PPIs at 1–1.7 mg/kg/day [118]. The esomeprazole dose of 10 mg/day proposed by the guideline for under 20 kg is, in an infant of 3 to 7.5 kg, higher than that of the US label, which grades the dose by weight and did not study it above 1.33 mg/kg/day; the guideline dose is the reference and the label is for comparison, but in the small infant the dose per kilogram should be checked, because the healing trials used 1–1.7 mg/kg/day [118, 119].

Sources of the table: [3, 17, 106, 108, 120].

7.11. Clinical trials and comparative studies

The evidence on the treatment of RS rests on single-centre, retrospective series with different definitions, populations and success criteria; the only randomised trial of a therapeutic intervention is the 48-hour crossover trial of PEBP versus sham plate, and the positioning trial showed no difference in the main analysis [93, 102]. It is not legitimate to compare indirectly the success rates in the tables for each intervention, because patients treated with positioning, NPA, PEBP or MDO differ in severity and syndromic status, and the PEBP was also applied to patients who would have been stable with positioning [3]. Resolution percentages should be read together with baseline severity: a 94% reduction in OAHI has a different meaning in a patient with 70 events per hour and in one with 10 [107, 112]. A search of ClinicalTrials.gov (16 records with "Robin", October 2026) found no pharmacological trials; the remaining records are observational or concern feeding, without results published in the registry, including one terminated and one withdrawn trial [120].

7.12. Developments in the last 12 months

No authorisations, regulatory opinions or pharmacological trials were identified in the last 12 months; this negative search does not prove the absence of ongoing studies [120].

7.13. Guideline recommendations and divergences

The following table gathers the divergences between the 2021 French PNDS and the other guidelines cited in this article [3, 4, 18, 26, 118].

The 2020 International Pediatric ORL Group (IPOG) consensus resulted from a two-round Delphi questionnaire, answered by 21 members from 16 tertiary centres in 8 countries (including Australia, Canada, France, Ireland, the United Kingdom, Belgium and the USA), and is based on expert opinion [26]. For assessment, it recommends clinical examination of breathing and feeding, flexible laryngoscopy in all (associated malformations, glossoptosis, difficulty of intubation, endoscopic signs of swallowing disorder), oximetry and PSG in the initial respiratory assessment, rigid laryngotracheobronchoscopy or flexible tracheobronchoscopy if an associated lesion is suspected (with airway CT as a complement or alternative), DISE for unexplained obstruction before or after surgery, craniofacial CT or cone-beam CT if there are associated craniofacial malformations or if MDO is being considered, and functional endoscopic evaluation of swallowing or videofluoroscopy for unexplained swallowing problems or aspiration; there was no consensus on the diagnosis of micrognathia [26]. For initial treatment, it proposes prone positioning as first line with in-hospital monitoring, the nasopharyngeal or oropharyngeal airway, CPAP (recommended especially when there is central apnoea) and intubation, or emergency tracheostomy if intubation fails, as a last resort [26]. For persistent obstruction, it considers CPAP (without discharge home if it is needed 24 hours a day, a situation in which other options should be weighed and CPAP maintained until surgery, with regular PSG to adjust the settings), prolonged NPA (UK and Canadian practice), TLA mainly with glossoptosis and moderate micrognathia, with unpredictable results, MDO mainly in severe micrognathia, and tracheostomy after failure of the first intervention or as first line in complex airway malformations or severe swallowing disorder; it recommends PSG after any procedure, found no consensus on the type of distractor or on the distraction protocol, and allows discharge when respiratory difficulty is controlled, with acceptable swallowing and adequate nutritional status [26]. In the prenatal period, it considers associated malformations, polyhydramnios and the severity of micrognathia most relevant for the perinatal plan, and proposes delivery in a tertiary centre when obstruction is anticipated, airway control (nasopharyngeal airway, laryngeal mask or intubation) on placental support during delivery and the EXIT procedure (ex utero intrapartum treatment) when very severe obstruction or difficult intubation are anticipated [26]. The differences from the European guideline are prone rather than lateral positioning as the first position, the proposal of EXIT and less emphasis on a systematic baseline sleep study [3, 26].

7.14. Situations to avoid, perioperative care, pregnancy and transition

Pharyngoplasty should be avoided whenever possible, because of the risk of inducing OSA or causing it to recur [3]. The combination of NGT and CPAP requires monitoring for pressure injuries, air leak and gastric distension [3]. Imaging with radiation at an early age should be weighed carefully, and DISE carries anaesthetic risks [3]. Before any anaesthetic, a difficult airway should be anticipated, previous intubation difficulties recorded and obstructions below the tongue base excluded [3, 95]. In type II collagenopathies and campomelic dysplasia, the cervical spine should be studied before anaesthesia and extreme flexion or extension of the neck avoided [34, 35]. After palatoplasty, MDO or TLA, respiratory and feeding monitoring is intensified [3, 116].

The sources consulted do not address pregnancy in women with RS; preconception counselling follows the aetiology (section 8). The European guideline recommends transition protocols to adulthood, and surveillance for OSA and occlusal problems continues until skeletal maturity, with orthognathic surgery in about 39% [3]. The French PNDS entrusts follow-up of the adult to adult specialists (otorhinolaryngology, maxillofacial surgery, orthodontics, speech and language therapy) and acknowledges that paediatric coordination is difficult to transfer: in isolated RS, the paediatrician passes the information on to the parents and the patient and tells them which specialists to consult; in non-isolated RS with a neurodevelopmental disorder, overall follow-up may be provided by the general practitioner, the neurologist, the geneticist or the doctor of the facility the adult attends [18]. The geneticist sees the patient again at the end of adolescence, to review the diagnosis in the light of molecular advances and discuss transmission before plans for parenthood [18]. The PNDS does not address pregnancy in women with RS.

8. Reproductive Options and Family Counselling

8.1. Principles

The recurrence risk depends on the cause of RS and not on the sequence itself, so counselling is only accurate after a complete aetiological assessment, which includes examination of the parents [2, 4]. Isolated CP is more often associated with malformations and syndromes than cleft lip and palate, and the parents should be examined for lip pits, bifid uvula and submucous cleft; in the presence of nasal speech, 22q11.2 deletion should be excluded [46]. Parents value the genetics clinic appointment at prenatal diagnosis and 6 to 12 months after birth, to discuss recurrence risks for the parents and, in the future, for the child; in a qualitative study, most parents did not remember having been seen by a geneticist [3].

8.2. Isolated RS without an identified cause

When RS is truly isolated, the recurrence risk is low, and the Brazilian textbook describes it as negligible in most sporadic cases, with rare families compatible with autosomal dominant inheritance with reduced penetrance and variable expressivity [2, 10]. The sources consulted do not provide empirical recurrence risks specific to isolated RS. In the Necker series, isolated RS included familial cases and a high incidence of twins, and in Bauru 27.7% of children with isolated RS had a family history of clefting [47, 71]. The French PNDS quantifies the familial fraction of isolated RS at 15% (10%–15% in the summary), a proportion of familial cases and not a recurrence risk [18]. In the follow-up by Sheffield and colleagues, there was no recurrence in siblings of patients with non-syndromic RS [6]. As an approximation, the Danish empirical risks for non-syndromic orofacial clefts in relatives of a proband with CP (more than 54,000 relatives, multifactorial threshold model) were 3.3% for siblings, 1.0% for half-siblings, 2.1% for parents, 4.2% for offspring, 1.1% for nephews and nieces, 0.6% for uncles and aunts and 0.4% for first cousins; these figures refer to non-syndromic CP in general and not to RS [46]. Before applying low risks, up-to-date genetic testing is needed, because reassessment changes the diagnosis in about a quarter of patients [3, 19].

Source of the table: Grosen and colleagues, Denmark, cited by the Oxford Desk Reference [46]; the column refers to the proband's phenotype and the risk to the occurrence of an orofacial cleft in relatives.

8.3. Lesions upstream of SOX9

Deletions and translocations of the regulatory region upstream of SOX9 are transmitted in an autosomal dominant manner, with possibly incomplete penetrance and variable expressivity: in a three-generation family, the same deletion produced RS, signs of acampomelic campomelic dysplasia and learning difficulties; a familial microdeletion was inherited from a mother with CP; familial translocations and deletions are described in the table in section 2.3 [9, 11, 13, 54]. A carrier parent has a 50% probability of transmitting the alteration to each child, but the offspring's phenotype cannot be predicted precisely; the phenotypic variability of SOX9 complicates prenatal counselling even when the alteration is known [7]. When the alteration is a balanced translocation, the parents' karyotype distinguishes a de novo origin from a familial translocation, and the reproductive risk of a carrier depends on the specific rearrangement [9]. In campomelic dysplasia due to coding SOX9 variants, penetrance is complete and recurrence due to gonadal mosaicism has been described [13, 78].

8.4. Syndromic forms

In dominant Stickler syndrome, each child of an affected person has a 50% probability of inheriting the variant [12]. In Catel-Manzke syndrome, which is autosomal recessive, heterozygous parents have a 25% probability of recurrence in each pregnancy [39]. In cerebro-costo-mandibular syndrome, most cases are sporadic, but familial transmission occurs and about 10% of families show non-penetrance [38, 60]. In SATB2-associated syndrome, usually de novo, recurrence in siblings due to parental germline mosaicism has been described, and when the proband has a 2q33.1 deletion or duplication the parents should have a karyotype to exclude a balanced translocation [15]. In Treacher Collins syndrome, expressivity is highly variable and the parent may be almost asymptomatic; once the variant has been identified, the parents should be tested to estimate the risk [2]. In TARP syndrome, which is X-linked recessive, carrier mothers are usually unaffected and each male child has a 50% probability of being affected; low-level mosaicism has been described in the mother of one patient, and a family history of miscarriages and of affected boys on the maternal side is a clue [41]. The risks of 22q11.2 deletion and of the other entities are specific to each.

8.5. Prenatal and preimplantation diagnosis

When the familial variant or rearrangement is known, prenatal diagnosis by amniocentesis or chorionic villus sampling and preimplantation genetic testing are possible; the phenotypic variability of lesions upstream of SOX9 should be discussed beforehand [3, 7]. Without a genetic diagnosis, ultrasound surveillance in a subsequent pregnancy looks for micrognathia and an abnormal profile from 18 weeks, with referral to a reference centre and, if available, MRI; CP may not be visible, and in Stickler syndrome absence of a cleft on the 19–20-week scan does not exclude the diagnosis [3, 12]. In a low-risk pregnancy in which a cleft is detected, the three-generation pedigree and assessment of relatives guide the investigation [12].

The French PNDS addresses termination of pregnancy for medical reasons in the case of prenatal suspicion of RS. The combination of several malformations, with or without a syndromic diagnosis, often has a poor prognosis and raises the question of termination if the parents request it; but in first arch syndromes and in many osteochondrodysplasias the intellectual prognosis is favourable, which requires collegial discussion at the CPDPN for any request [18]. If the parents opt for termination, fetopathological examination is indispensable and may provide the diagnosis and genetic counselling for a subsequent pregnancy [18]. In isolated fetal retrognathia, the child may be healthy or have isolated RS or RS associated with minor anomalies, especially bony ones, that only become apparent later; since cognitive development in isolated RS, or RS associated with bone dysplasias, is normal and morbidity and mortality are low, the PNDS considers that this situation does not justify termination [18]. The legal framework for termination of pregnancy for medical reasons is the French one and cannot be transposed directly to other countries.

8.6. Cascade testing and counselling of minors

Once the cause has been identified, at-risk relatives can be tested; in dominant syndromes with variable expressivity, clinical examination and testing of apparently unaffected parents refine the risk [2, 15]. Testing minors is justified when the result changes surveillance in childhood, as in Stickler syndrome, in which ophthalmological examination and prevention of retinal detachment depend on the diagnosis [12]. Counselling of the patient with isolated RS should be repeated in adolescence or adulthood, when the recurrence risks for their offspring become relevant [3]. The French PNDS considers genetic counselling useful in all situations and provides for the geneticist to see the adolescent again, even with isolated RS, because exome and genome sequencing may provide new leads, and to review the diagnosis with the young adult before plans for parenthood; in the syndromic form, genetic follow-up answers questions about the risk of transmission, and in the associated form without a diagnosis it continues the investigation with the available techniques [18].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Birth prevalence

The meta-analysis by Wright and colleagues (34 studies, 2722 cases, 22 countries, 53% European) estimated a birth prevalence of 9.5 per 100,000 live births (95% CI 7.1–12.1), about 1 in 10,500, with extreme heterogeneity (I² 98%) and a range of 1.2 to 40.4 per 100,000 [17]. The European estimate was 10.4 per 100,000 (95% CI 7.3–14.0; 18 studies) and the non-European 8.3 (95% CI 4.9–12.7; 16 studies), with no significant difference; registries gave 8.6 and non-registry studies 10.4 [17]. An earlier review, up to 2012, gave a median of 1 in 14,500 (6.5 per 100,000) and a range of 1 in 3900 to 1 in 122,400 [17, 125]. The European guideline gives 1 in 8500 to 1 in 14,000, with low reliability because of variable definitions, and the 2016 consensus about 1 in 8500 [3, 4]. The Oxford Desk Reference gives 1 in 8500, and the Brazilian textbook of syndromes usually gives 1 in 8500, with a range of 1 in 2000 to 1 in 30,000 and equal frequency in both sexes [10, 46]. The 2021 French PNDS gives an incidence of 1 in 10,000 to 1 in 14,000 births, depending on definitions and studies, and cites two lower European estimates, both limited to RS with CP and with few cases: 1 in 30,827 births in the 20 registries from 12 countries analysed in 2003 (23 RS in 709,030 fetuses) and the Norwegian series of 49,314 births (1987–2004), in which posterior CP occurred in 1 in 2054 [18].

Orphanet records, for isolated RS, a European expert estimate of 5.0 per 100,000 and validated national values of 8.0 in France, 12.4 in Germany, 3.5 in Denmark, 4.5 in Sweden, 5.9 in the United Kingdom, 6.4 in the Netherlands and 18.8 in Australia, and a non-validated value of 7.9 in the USA [5]. The Orphanet values refer to isolated RS, whereas most population studies count all RS, which explains part of the differences.

9.2. What explains the discrepancies

Estimates vary more than 30-fold for methodological reasons, which is why the French PNDS recommends recording "RS with cleft" and "RS without cleft" separately in databases [18]. Only 11 of the 34 studies in the meta-analysis stated a definition, and 23 (68%) included only RS with CP, which underestimates prevalence, because RS without CP accounts for 8%–17% of cases [17]. Cleft registry studies do not capture RS without a cleft; administrative databases depend on coding, and the figure of 1 in 1940 from the American HCUP-KID database, well above the other estimates, should be read with this limitation in mind [25]. The Dutch study by Paes and colleagues illustrates the difficulty: the original abstract gives 1 in 5600 live births (17.7 per 100,000), Smith's textbook cites the same figure, the meta-analysis recalculated 246 cases in 2,113,858 births (11.6 per 100,000) and Orphanet records 6.4 per 100,000 for isolated RS; the sources consulted do not allow these figures to be fully reconciled, and they probably differ in the denominator and in the inclusion of syndromic forms [1, 5, 17, 125]. The same study showed that RS accounts for about one third of the population with CP, with more severe clefts, and, in the review of 42 studies, a range of 0.8 to 32.0 per 100,000 and a mean of 8.0 per 100,000 [125]. For reference, isolated CP occurs in about 4 per 10,000 births [46].

9.3. Sex, ethnicity and proportion of isolated forms

In the estimates for total births, 52% of cases were female; the Nijmegen cohort had 61% girls and the Maltese cohort 63.6% boys, which shows variation between small series [17, 20, 127]. In California, prevalence was higher among children of non-Hispanic white women (8 per 100,000) and lower among children of Asian women (1.7); in the American multistate study, it was 17–21 per 100,000 in white and 5–11 in black infants, with a predominance of the isolated form in white and of the non-isolated form in black infants [17]. Isolated RS accounted for 62% in 7 population studies and 68% in the European congenital anomaly registry EUROCAT [17].

9.4. Populations and national data

Europe. Germany, the Netherlands, France and the United Kingdom have population studies and reference cohorts (Tübingen, Utrecht, Necker, Cleft Collective) [17, 22]. Practices vary: the PEBP is first line in Germany, the NPA in Scotland, and tracheostomy was the most used surgical intervention in two European studies [17]. In France, the 2021 PNDS describes prone positioning with EN as the first line for all teams, NIV and NPA as second line and tracheostomy as the technique for relieving obstruction most used until then in severe RS, and notes that TLA is only practised in Lille, Marseille and Tours [18].

United States. Population estimates vary between 5.4 and 32 per 100,000 depending on the period and method; the USA treats most patients surgically (77% in a multistate study) and prefers MDO [17, 115].

Brazil. The Hospital for Rehabilitation of Craniofacial Anomalies of the University of São Paulo (HRAC-USP, "Centrinho"), in Bauru, developed a single protocol for all RS, based on nasopharyngoscopy, nasopharyngeal intubation and feeding techniques, and produced studies of natural history, feeding, reflux, speech, craniofacial morphology, dentition and sleep [3, 47, 70, 101]. The only Brazilian prevalence estimate included in the meta-analysis comes from Campos dos Goytacazes (12.8 per 100,000, cleft only, 6 cases) [17]. The clinical case in section 10.1 comes from São Paulo [9]. One of the authors of the 2016 international consensus represented the Hospital de Clínicas de Porto Alegre [4].

Portugal. The sources consulted do not include Portuguese epidemiological data; the Brazilian textbook cites a 2015 thesis from the Faculty of Dental Medicine of the University of Lisbon and a 2007 article in the Revista Portuguesa de Estomatologia, Medicina Dentária e Cirurgia Maxilofacial, without prevalence figures [10]. The lack of data in the sources consulted does not demonstrate their non-existence.

Founder populations. The TGDS variant c.298G>T p.(Ala100Ser), present on at least one allele in most patients with Catel-Manzke syndrome, is considered a probable European founder variant [39]. No founder variants have been described for isolated RS.

9.5. Special populations

Low birth weight and neurological compromise are associated with failure of MDO and the need for tracheostomy; neurological anomaly is associated with higher mortality [3, 24]. In a series of 121 MDOs, infants weighing less than 4 kg had success (92.6% versus 88.9%), site infection (9.9% versus 20.0%) and complications (17.3% versus 25.0%) similar to or lower than those of infants weighing more than 4 kg, with 2 major complications in the lower-weight group [3]. Children with RS and CP have an increased risk of hearing loss, and those who had prolonged respiratory support have more persistent sleep disorders [3, 23]. Unequal access to prolonged admission and to devices such as the PEBP may lead to surgery in children who would be treatable without it [3, 106].

9.6. European network and patient organisations

ERN CRANIO (the European Reference Network for rare craniofacial anomalies and ear, nose and throat disorders, coordinated by Erasmus MC in Rotterdam) is the relevant network and produced the European guideline; it includes pages for patients, patient representatives (European Patient Advocacy Groups, ePAG) and a registry [3, 28]. Stichting Pierre Robin Europe, based in Amsterdam, is the organisation of patients and clinicians that represents RS in ERN CRANIO [29]. In France, the SPRATON reference centre (coordinating site at Necker, Paris, and 21 competence centres accredited in 2017) belongs to the TETECOU rare disease health network and publishes on its website (https://www.tete-cou.fr/spraton/accueil) the directory of centres, the care pathway, the prone positioning sheet, information for maternity units, an information booklet for families, practical sheets and an emergency card [18, 105]. The association Tremplin – Syndrome de Pierre Robin (https://www.tremplin-spr.org), founded by parents more than 20 years ago and integrated into the TETECOU network, offers listening and support, puts families in contact and provides information on the disease, the care network and social services, and organises annual meetings between professionals and families; the PNDS recommends informing families of its existence as early as possible [18]. On the social side, the PNDS considers that all children with RS justify 100% coverage by health insurance (as a condition outside the list of chronic diseases), that infants who return home with technical care (EN, NIV, NPA) justify the education allowance for a disabled child (AEEH), because one of the parents generally has to stop working or pay for home childcare, and that, without technical care, the daily parental presence allowance (AJPP) can be requested; group childcare is discouraged, or even contraindicated, in the first year, to avoid ear, nose and throat infections that worsen the obstruction [18]. After the first year, the AEEH is only justified in non-isolated RS with disability or expenses for prostheses (hearing aids, glasses); children with isolated RS do not need school support, and those with syndromic RS receive the support that the syndrome justifies (assistant for pupils with disabilities, personalised schooling plan) [18]. The PNDS also refers to the Orphanet page on isolated RS and to the Orphanet encyclopaedia article for the public, and to the rare disease platform of AP-HP Centre-Université de Paris [18]. Other organisations cited by the sources are the Cleft Lip and Palate Association (CLAPA, United Kingdom; https://www.clapa.com/), the American Cleft Palate-Craniofacial Association (ACPA, USA; https://acpacares.org/) and Contact a Family, in the United Kingdom [2]. HRAC-USP (https://hrac.usp.br/) is the Brazilian public reference centre. The Pierre Robin Network domain, cited in earlier editions of the textbooks, was not responding in October 2026. The European guideline recommends uniform online information in the parents' language, because about 80% search for information on the internet, which is sometimes non-existent in their mother tongue or generates anxiety [3].

10. Illustrative Clinical Case

10.1. RS with a balanced translocation upstream of SOX9 (São Paulo, Brazil)

Fonseca and colleagues, of the University of São Paulo, described in 2013 a boy, the second child of healthy non-consanguineous parents, born at term after an uneventful pregnancy, with a birth weight of 3350 g (50th centile), length of 47 cm (10th centile), head circumference of 36 cm (50th centile) and Apgar scores of 7, 8 and 10 [9]. Posterior CP and micro-retrognathia led to the hypothesis of RS, and a chest radiograph showed scoliosis. He had respiratory difficulty treated with oxygen and was discharged at 50 days; two weeks later he was readmitted with respiratory difficulty, pallor, apnoea and severe bradycardia, which required oxygen and chest compressions, followed by mechanical ventilation in the intensive care unit. Weaning was difficult, and at 6 months he underwent tracheostomy and gastrostomy; tracheobronchomalacia was diagnosed and ventilation was prolonged. He sat at 18 months and walked at 30 months [9].

At 3 years he weighed 14 kg (25th–50th centile), measured 91 cm (10th–25th centile) and had a head circumference of 51 cm (50th–98th centile); he had a flat face, bilateral epicanthus, low nasal bridge, anteverted nares, posterior CP, mild micrognathia, short neck, hypoplastic nipples and severe kyphoscoliosis, with normal male genitalia and proportionate limbs [9]. Radiographs showed severe thoracic kyphoscoliosis, non-mineralised thoracic vertebral pedicles, narrow iliac wings, delayed ischiopubic ossification, a flattened femoral epiphysis with an elongated neck, absent lesser trochanter and a short fibula, which led to the diagnosis of acampomelic campomelic dysplasia (OMIM 114290) [9]. The karyotype revealed an apparently balanced de novo translocation t(17;20)(q24.3;q11.2), with normal parental karyotypes; sequencing of the coding region of SOX9 and the exon-intron junctions was normal, and the microarray (Agilent 105A) showed no imbalances [9]. Fluorescence in situ hybridisation mapping located the 17q breakpoint in a segment of about 16 kb (chr17:69,516,104-69,531,685, hg19), in a gene desert 585–601 kb upstream of SOX9, and the 20q11.2 breakpoint in a segment of about 50 kb, with possible disruption of C20orf4 [9]. Figure 2 of the article shows the in situ hybridisation mapping of the translocation (https://pmc.ncbi.nlm.nih.gov/articles/PMC3658899/figure/F2/) and Figure 3 the diagram of the rearrangements upstream of SOX9, with the proximal, distal and RS clusters (https://pmc.ncbi.nlm.nih.gov/articles/PMC3658899/figure/F3/) [9].

In the same article, a second patient with a de novo t(7;17)(p13;q24) and a breakpoint 855–917 kb upstream of SOX9 (chr17:69,201,539-69,262,086, hg19) had repeated bronchitis, short stature at 12 years (136.5 cm, 3rd centile), hypoplastic scapulae, thoracolumbar scoliosis and 11 pairs of ribs, with normal intellectual development, without RS described; at 31 years, at 159 cm, she sought counselling on the risk to her offspring (Figure 1: https://pmc.ncbi.nlm.nih.gov/articles/PMC3658899/figure/F1/) [9]. Based on these patients, the authors redefined the distal breakpoint cluster down to 585–601 kb and noted that carriers of balanced rearrangements live longer than carriers of SOX9 point variants [9].

The case illustrates four points. First, the neonatal presentation as apparently isolated RS may conceal a skeletal dysplasia that only becomes evident with growth, which justifies skeletal radiographs and genetic reassessment [3, 4]. Second, SOX9 sequencing and microarray were normal, and only the karyotype identified the cause, an example of the limits of the first-line tests described in section 5.6 [3, 9]. Third, the respiratory deterioration two weeks after discharge occurred within the window of late obstruction of the first months; current recommendations, which postdate the case, provide for baseline and repeated sleep studies, FNE to identify all levels of obstruction (here, tracheobronchomalacia) and surveillance for CO2 retention during the first month, which do not appear in the report [3]. Fourth, the management reported (discharge without a described sleep study, tracheostomy at 6 months) reflects the practice of the centre and the period and should not be taken as routine.

10.2. Difficult neonatal airway with a tongue-base lesion (Nanjing, China)

Zhang and colleagues described a 9-day-old boy, weighing 3.5 kg, with RS, complete unilateral cleft palate (maximum width of about 0.8 cm), atrial septal defect and aspiration pneumonia, with cyanosis and venous CO2 pressure (PvCO2) of 87.8 mmHg [95]. Direct laryngoscopy and videolaryngoscopy failed (Cormack-Lehane grade 4), and the size 1 laryngeal mask was blocked by a tongue-base lesion; the NPA restored an SpO2 of 100% [95]. On day 3, in theatre for MDO, under sevoflurane with spontaneous ventilation, fibreoptic examination showed the epiglottis lying against the posterior wall; the laryngeal mask and lighted stylet failed and the procedure was abandoned; with a PvCO2 of 119.4 mmHg, TLA was performed under sedation, but the following day the PvCO2 reached 183.8 mmHg [95]. Review of the CT identified a 21.1 × 11.7 mm lesion at the tongue base, described as a vallecular cyst in the abstract and as a probable thyroglossal duct cyst by a second radiologist [95]. On day 5, after induction with ketamine and sevoflurane, intubation was achieved guided by the middle finger, which lifted the epiglottis, with a tube curved according to the three-dimensional reconstruction; the initial end-tidal CO2 (ETCO2) was 120 mmHg, MDO proceeded without problems and extubation took place on the 5th postoperative day [95]. Figure 1 compares the patient's craniofacial CT reconstruction with that of a 2-month-old with RS (https://pmc.ncbi.nlm.nih.gov/articles/PMC6530166/figure/Fig1/) and Figure 2 shows the airway CT with the lesion (https://pmc.ncbi.nlm.nih.gov/articles/PMC6530166/figure/Fig2/) [95]. The centre treated 225 RS in 2017 [95].

The case shows that the obstruction of an infant with RS may be multilevel and include lesions unrelated to the micrognathia, and that failure of the usual intubation techniques is possible. The European guideline recommends endoscopy of the whole airway before any surgery and initial non-surgical treatment; digital intubation is a rescue technique described in a single report and not routine practice [3, 95].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The prognosis depends mainly on the cause and on the severity of the obstruction. In isolated RS, it is good if the child survives the initial period of obstruction, with mortality of 0%–2.2% in recent series; in syndromic RS, mortality was 14.1%–15% and was associated with neurological and cardiac anomalies [1, 21, 24, 25]. The Brazilian textbook considers the cases that needed tracheostomy to be more severe [10]. In cerebro-costo-mandibular syndrome, historical mortality was about 25% in the neonatal period and 56% at 1 year, and about 20% with current intensive care; the number of rib defects predicts lethality, and after the first year respiratory difficulties result from progressive scoliosis [60]. In terminal 4q deletion, 50% died before 15 months [62]. In Catel-Manzke syndrome, growth may normalise if the child survives early infancy, and most have normal intelligence [39, 99].

OSA improves in most during the first year: the median OAHI falls from 17 to 5 per hour between 2 and 8 months [64]. The NPA is removed after weeks to months, TLA is taken down at about 6–8 months and decannulation after tracheostomy occurred at a mean age of 3 years in one series [3]. OSA may, however, persist or reappear: at school age, a child without complaints had severe OSA, and in Bauru OSA screening was positive in about 60% of RS aged 6 to 12 years [69, 70]. Weight gain recovers in most during the first year, with a weight z score of −0.44 at one year in a large cohort treated with PEBP [3]. The mandible retains a sagittal deficit, and about 39% need orthognathic surgery at maturity [3]. Neurodevelopmental deficit affects about 10% in isolated RS and 35% in non-isolated RS [68]. The feeding difficulties of severe isolated RS resolve in most: in the prospective Necker series, they persisted in 18% at 3 years and in 6% at 6 years [18]. The French PNDS stresses that a child with RS should have normal height and weight growth, unless RS is part of a syndrome that affects growth [18]. The quality of life of school-age children and adolescents is similar to the norms in most domains, with more sleep and physical functioning problems reported by parents in childhood and high parental distress in the first year [3, 117].

11.2. Surveillance

The frequency of the dental appointment is ambiguous in the original guideline ("bi-annual"), and that of speech assessments differs between the summary and the chapter of the guideline (section 7.6) [3]. The French PNDS also has internal differences between the summary and the text (age at the maxillofacial appointment, audiological schedule), and the more detailed intervals of the text are those given in the table; the document clarifies that these intervals, without a published basis, result from the experience of the reference centre and from families' requests and represent the useful minimum in isolated RS, being at least the same in syndromic RS, in which confirmation of the diagnosis is awaited until the first year [18]. The syndromic forms add the surveillance specific to each syndrome.

12. Summary and Concluding Remarks

Robin sequence is the cascade of micrognathia, glossoptosis and upper airway obstruction, with a U-shaped CP in most patients, resulting from early restriction of mandibular development. It is not an aetiological diagnosis: fewer than half of patients have isolated RS after genetic reassessment, and Stickler syndrome is the most frequent syndromic cause, followed, depending on the series, by 22q11.2 deletion, Treacher Collins syndrome and rare entities such as Catel-Manzke, cerebro-costo-mandibular, Richieri-Costa-Pereira, SATB2-associated and TARP syndromes. In isolated RS, the best-established cause is loss of neural crest enhancers located more than 1 Mb upstream of SOX9, at 17q24.3, a non-coding lesion that escapes exome sequencing and requires high-density microarray, karyotyping or genome sequencing.

The 2016 international consensus and the 2024 ERN CRANIO European guideline established a common definition (primary micrognathia, mandatory glossoptosis and obstruction, additional CP) and an assessment based on sleep studies, endoscopy of the whole airway, feeding assessment and genetic testing with CNV analysis and an exome panel, with reassessment at 2–3 years. Clinical assessment underestimates the severity of OSA, which may develop up to 2 months of age. Treatment escalates from positioning, the nasopharyngeal airway, the pre-epiglottic baton plate and CPAP to mandibular distraction osteogenesis, tongue-lip adhesion and tracheostomy, with no demonstrated superiority of any strategy and with mainly observational evidence; the safety of prone positioning, access to the pre-epiglottic baton plate and the timing of surgery are the main points of divergence between centres. The 2021 French PNDS adds a national view that differs on several points: it keeps CP in the definition, favours home prone positioning with strict rules, NIV and NPA, almost completely abandons TLA, reserves early MDO for major mandibulofacial dysostoses, sets intervals for paediatric, otorhinolaryngological, maxillofacial, orthodontic and speech and language therapy follow-up and organises care in the SPRATON reference centre. Follow-up extends to feeding and growth, palatoplasty with sleep studies before and after, speech, hearing, dentition with agenesis in more than 40%, mandibular growth that does not fully catch up and psychosocial support for the family. The prognosis is good in treated isolated RS and depends, in the syndromic forms, on the underlying disease; genetic counselling is only accurate once the aetiology has been clarified.

13. References

1. Jones KL, Jones MC, del Campo M. Robin sequence (Pierre Robin syndrome). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 334–5.
2. Firth HV, Hurst JA. Micrognathia and Robin sequence. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 232–5.
3. By the Working Group on Writing a European Guideline on Robin Sequence. European Guideline Robin Sequence An Initiative From the European Reference Network for Rare Craniofacial Anomalies and Ear, Nose and Throat Disorders (ERN-CRANIO). J Craniofac Surg. 2024;35(1):279-361. doi:10.1097/SCS.0000000000009701. PMID: 37811988.
4. Breugem CC, Evans KN, Poets CF, Suri S, Picard A, Filip C, et al. Best Practices for the Diagnosis and Evaluation of Infants With Robin Sequence: A Clinical Consensus Report. JAMA Pediatr. 2016;170(9):894-902. doi:10.1001/jamapediatrics.2016.0796. PMID: 27429161.
5. Orphanet. Isolated Pierre Robin sequence (ORPHA:718). Available from: https://www.orpha.net/en/disease/detail/718
6. Online Mendelian Inheritance in Man, OMIM®. MIM Number 261800: Pierre Robin syndrome; PRBNS. Baltimore (MD): Johns Hopkins University. Accessed 2026 Oct 10. Available from: https://omim.org/entry/261800
7. van de Velde S, Mink van der Molen AB, Lachmeijer AMA, de Leijer D, Smits JJ, Massink MPG, et al. Genetic Landscape of Robin Sequence: A Systematic Review. Clin Genet. 2026;109(2):218-232. doi:10.1111/cge.70088. PMID: 41077824. PMC12779227.
8. Benko S, Fantes JA, Amiel J, Kleinjan DJ, Thomas S, Ramsay J, et al. Highly conserved non-coding elements on either side of SOX9 associated with Pierre Robin sequence. Nat Genet. 2009;41(3):359-64. doi:10.1038/ng.329. PMID: 19234473.
9. Fonseca AC, Bonaldi A, Bertola DR, Kim CA, Otto PA, Vianna-Morgante AM. The clinical impact of chromosomal rearrangements with breakpoints upstream of the SOX9 gene: two novel de novo balanced translocations associated with acampomelic campomelic dysplasia. BMC Med Genet. 2013;14:50. doi:10.1186/1471-2350-14-50. PMID: 23648064. PMC3658899.
10. Rocha Júnior JLG, Ribeiro EM. Sequência de Pierre Robin (SPR). In: Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 197–9.
11. Castori M, Bottillo I, Morlino S, Barone C, Cascone P, Grammatico P, et al. Variability in a three-generation family with Pierre Robin sequence, acampomelic campomelic dysplasia, and intellectual disability due to a novel ∼1 Mb deletion upstream of SOX9, and including KCNJ2 and KCNJ16. Birth Defects Res A Clin Mol Teratol. 2016;106(1):61-8. doi:10.1002/bdra.23463. PMID: 26663529.
12. Mortier G. Stickler syndrome. 2000 Jun 9 [updated 2023 Sep 7]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1302/ PMID: 20301479.
13. Unger S, Scherer G, Superti-Furga A. Campomelic dysplasia. 2008 Jul 31 [updated 2021 Mar 18; revised 2023 Apr 6]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1760/ PMID: 20301724.
14. McDonald-McGinn DM, Hain HS, Emanuel BS, Zackai EH. 22q11.2 deletion syndrome. 1999 Sep 23 [updated 2020 Feb 27; revised 2025 May 8]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1523/ PMID: 20301696.
15. Zarate YA, Bosanko K, Fish J. SATB2-associated syndrome. 2017 Oct 12 [updated 2024 Jun 20]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK458647/ PMID: 29023086.
16. Gregersen PA, Savarirayan R. Type II collagen disorders overview. 2019 Apr 25 [updated 2024 Oct 24]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK540447/ PMID: 31021589.
17. Wright M, Cortina-Borja M, Knowles R, Urquhart DS. Global birth prevalence of Robin sequence in live-born infants: a systematic review and meta-analysis. Eur Respir Rev. 2023;32(170). doi:10.1183/16000617.0133-2023. PMID: 38056889. PMC10698549.
18. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Séquence de Pierre Robin. Centre de Référence Maladies Rares des Syndromes de Pierre Robin et troubles de succion-déglutition congénitaux (SPRATON), Filière de Santé Maladies Rares TETECOU; texte de septembre 2021. Saint-Denis La Plaine: HAS; 2021 Oct 20. Available from: https://www.has-sante.fr/jcms/p_3293455
19. Basart H, Paes EC, Maas SM, van den Boogaard MJ, van Hagen JM, Breugem CC, et al. Etiology and pathogenesis of robin sequence in a large Dutch cohort. Am J Med Genet A. 2015;167A(9):1983-92. doi:10.1002/ajmg.a.37154. PMID: 26033782.
20. van den Elzen AP, Semmekrot BA, Bongers EM, Huygen PL, Marres HA. Diagnosis and treatment of the Pierre Robin sequence: results of a retrospective clinical study and review of the literature. Eur J Pediatr. 2001;160(1):47-53. doi:10.1007/s004310000646. PMID: 11195018.
21. Costa MA, Tu MM, Murage KP, Tholpady SS, Engle WA, Flores RL. Robin sequence: mortality, causes of death, and clinical outcomes. Plast Reconstr Surg. 2014;134(4):738-745. doi:10.1097/PRS.0000000000000510. PMID: 25357033.
22. Davies A, Davies A, Wren Y, Deacon S, Cobb A, McLean N, et al. Syndromes associated with Robin sequence: a national prospective cohort study. Arch Dis Child. 2023;108(1):42-46. doi:10.1136/archdischild-2022-324722. PMID: 36376018.
23. Silva BCS, Silva JM, Brandão GR, Alvarenga KF. Audiological Findings in Robin Sequence: A Systematic Review and Meta-Analysis. Cleft Palate Craniofac J. 2026;:10556656261488464. doi:10.1177/10556656261488464. PMID: 42736632.
24. Logjes RJH, Haasnoot M, Lemmers PMA, Nicolaije MFA, van den Boogaard MH, Mink van der Molen AB, et al. Mortality in Robin sequence: identification of risk factors. Eur J Pediatr. 2018;177(5):781-789. doi:10.1007/s00431-018-3111-4. PMID: 29492661. PMC5899115.
25. Chocron Y, Alabdulkarim A, Czuzoj-Shulman N, Abenhaim HA, Gilardino MS. Incidence, Characteristics, and Outcomes of Robin Sequence: A Population-Based Analysis in the United States. J Craniofac Surg. 2024;35(1):6-9. doi:10.1097/SCS.0000000000009649. PMID: 37622565.
26. Fayoux P, Daniel SJ, Allen G, Balakrishnan K, Boudewyns A, Cheng A, et al. International Pediatric ORL Group (IPOG) Robin Sequence consensus recommendations. Int J Pediatr Otorhinolaryngol. 2020;130:109855. doi:10.1016/j.ijporl.2019.109855. PMID: 31896499.
27. Savarirayan R, Bompadre V, Bober MB, Cho TJ, Goldberg MJ, Hoover-Fong J, et al. Best practice guidelines regarding diagnosis and management of patients with type II collagen disorders. Genet Med. 2019;21(9):2070-2080. doi:10.1038/s41436-019-0446-9. PMID: 30696995.
28. European Reference Network for rare and/or complex craniofacial anomalies and ENT disorders (ERN CRANIO). Available from: https://www.ern-cranio.eu/
29. National Organization for Rare Disorders. Stichting Pierre Robin Europe [patient organisation]. Available from: https://rarediseases.org/organizations/stichting-pierre-robin-europe
30. Orphanet. Pierre-Robin sequence-Manzke dysostosis-clinodactyly of the index finger syndrome (ORPHA:1388). Available from: https://www.orpha.net/en/disease/detail/1388
31. Orphanet. Cerebrocostomandibular syndrome (ORPHA:1393). Available from: https://www.orpha.net/en/disease/detail/1393
32. Orphanet. Richieri Costa-Pereira syndrome (ORPHA:3102). Available from: https://www.orpha.net/en/disease/detail/3102
33. Orphanet. Campomelic dysplasia (ORPHA:140). Available from: https://www.orpha.net/en/disease/detail/140
34. GeneGuidance. SOX9 — clinical gene curation. Local project database. Accessed 2026 Oct 10.
35. GeneGuidance. COL2A1 — clinical gene curation. Local project database. Accessed 2026 Oct 10.
36. GeneGuidance. COL11A1 — clinical gene curation. Local project database. Accessed 2026 Oct 10.
37. GeneGuidance. TGDS — clinical gene curation. Local project database. Accessed 2026 Oct 10.
38. GeneGuidance. SNRPB — clinical gene curation. Local project database. Accessed 2026 Oct 10.
39. Jones KL, Jones MC, del Campo M. Catel-Manzke syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 414–5.
40. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-8. doi:10.1002/humu.23348. Available from: https://variantvalidator.org/
41. Online Mendelian Inheritance in Man, OMIM®. MIM Number 311900: TARP syndrome; TARPS. Baltimore (MD): Johns Hopkins University. Accessed 2026 Oct 10. Available from: https://omim.org/entry/311900
42. Firth HV, Hurst JA. Approach to the consultation: dysmorphism and developmental delay. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 4–7.
43. Jamshidi N, Macciocca I, Dargaville PA, Thomas P, Kilpatrick N, McKinlay Gardner RJ, et al. Isolated Robin sequence associated with a balanced t(2;17) chromosomal translocation. J Med Genet. 2004;41(1):e1. doi:10.1136/jmg.2003.010157. PMID: 14729841. PMC1757243.
44. Jakobsen LP, Ullmann R, Christensen SB, Jensen KE, Mølsted K, Henriksen KF, et al. Pierre Robin sequence may be caused by dysregulation of SOX9 and KCNJ2. J Med Genet. 2007;44(6):381-6. doi:10.1136/jmg.2006.046177. PMID: 17551083. PMC2740883.
45. Long HK, Osterwalder M, Welsh IC, Hansen K, Davies JOJ, Liu YE, et al. Loss of Extreme Long-Range Enhancers in Human Neural Crest Drives a Craniofacial Disorder. Cell Stem Cell. 2020;27(5):765-783.e14. doi:10.1016/j.stem.2020.09.001. PMID: 32991838. PMC7655526.
46. Firth HV, Hurst JA. Cleft lip and palate. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 102–5.
47. Marques IL, Barbieri MA, Bettiol H. Etiopathogenesis of isolated Robin sequence. Cleft Palate Craniofac J. 1998;35(6):517-25. doi:10.1597/1545-1569_1998_035_0517_eoirs_2.3.co_2. PMID: 9832223.
48. Logjes RJH, Breugem CC, Van Haaften G, Paes EC, Sperber GH, van den Boogaard MH, et al. The ontogeny of Robin sequence. Am J Med Genet A. 2018;176(6):1349-1368. doi:10.1002/ajmg.a.38718. PMID: 29696787.
49. Jones KL, Jones MC, del Campo M. Moebius sequence. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 322–3.
50. Gordon CT, Tan TY, Benko S, Fitzpatrick D, Lyonnet S, Farlie PG. Long-range regulation at the SOX9 locus in development and disease. J Med Genet. 2009;46(10):649-56. doi:10.1136/jmg.2009.068361. PMID: 19473998.
51. Gordon CT, Attanasio C, Bhatia S, Benko S, Ansari M, Tan TY, et al. Identification of novel craniofacial regulatory domains located far upstream of SOX9 and disrupted in Pierre Robin sequence. Hum Mutat. 2014;35(8):1011-20. doi:10.1002/humu.22606. PMID: 24934569. PMC4389788.
52. Fukami M, Tsuchiya T, Takada S, Kanbara A, Asahara H, Igarashi A, et al. Complex genomic rearrangement in the SOX9 5' region in a patient with Pierre Robin sequence and hypoplastic left scapula. Am J Med Genet A. 2012;158A(7):1529-34. doi:10.1002/ajmg.a.35308. PMID: 22529047.
53. Smyk M, Roeder E, Cheung SW, Szafranski P, Stankiewicz P. A de novo 1.58 Mb deletion, including MAP2K6 and mapping 1.28 Mb upstream to SOX9, identified in a patient with Pierre Robin sequence and osteopenia with multiple fractures. Am J Med Genet A. 2015;167A(8):1842-50. doi:10.1002/ajmg.a.37057. PMID: 26059046.
54. Amarillo IE, Dipple KM, Quintero-Rivera F. Familial microdeletion of 17q24.3 upstream of SOX9 is associated with isolated Pierre Robin sequence due to position effect. Am J Med Genet A. 2013;161A(5):1167-72. doi:10.1002/ajmg.a.35847. PMID: 23532965.
55. Uttley K, Jüllig HJ, De Angelis C, Auer JMT, Ozga E, Bengani H, et al. Neanderthal-derived variants increase SOX9 enhancer activity in craniofacial progenitors that shape jaw development. Development. 2025;152(21). doi:10.1242/dev.204779. PMID: 41208708. PMC12669974.
56. Xu JX, Kilpatrick N, Baker NL, Penington A, Farlie PG, Tan TY. Clinical and Molecular Characterisation of Children with Pierre Robin Sequence and Additional Anomalies. Mol Syndromol. 2016;7(6):322-328. doi:10.1159/000449115. PMID: 27920635. PMC5131331.
57. Jones KL, Jones MC, del Campo M. Stickler syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 410–3.
58. Jones KL, Jones MC, del Campo M. 22q11.2 (DiGeorge/velocardiofacial) syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 386–91.
59. Jones KL, Jones MC, del Campo M. Treacher Collins syndrome (mandibulofacial dysostosis). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 362–3.
60. Jones KL, Jones MC, del Campo M. Cerebro-costo-mandibular syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 844–7.
61. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 41, 41.1 Glycosaminoglycans synthesis defects.
62. Jones KL, Jones MC, del Campo M. Deletion 4q syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 34–5.
63. Jones KL, Jones MC, del Campo M. 15q24 microdeletion syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 90–1.
64. Coutier L, Gyapay R, Guyon A, Thieux M, Pouyau R, Blanc S, et al. Role of polysomnography in the management of obstructive sleep apnea during the first year of life in robin sequence: A prospective and longitudinal study. Sleep Med. 2025;126:73-81. doi:10.1016/j.sleep.2024.11.041. PMID: 39642651.
65. Wiechers C, Sowula J, Kreutzer K, Schwarz CE, Weismann C, Krimmel M, et al. Prospective cohort study on facial profile changes in infants with Robin sequence and healthy controls. World J Pediatr. 2024;20(6):581-589. doi:10.1007/s12519-024-00797-z. PMID: 38578561. PMC11239728.
66. Schafer S, Yu J, Mulliken JB. Catch-Up Growth in Syndromic Robin Sequence: A Systematic Review. Cleft Palate Craniofac J. 2026;63(10):3258-3270. doi:10.1177/10556656251392267. PMID: 41252330.
67. Klintö K, Schaar Johansson M, Brunnegård K, Becker M. Speech Outcomes in 5-Year-Olds Born With Cleft Palate With and Without Robin Sequence-A Swedish Registry Study. Cleft Palate Craniofac J. 2026;63(10):3247-3257. doi:10.1177/10556656251387538. PMID: 41124312. PMC13542379.
68. Goel D, Wilson A, Baynam G, Waters K, Pillow J, Rao S. Neurodevelopmental impairment in children with Robin sequence: A systematic review and meta-analysis. Early Hum Dev. 2025;201:106185. doi:10.1016/j.earlhumdev.2024.106185. PMID: 39733593.
69. Wiechers C, Uhlig S, Poets A, Weise C, Naros A, Krimmel M, et al. Sleep and neurocognitive outcome in primary school children with Robin Sequence. Sleep. 2023;46(5). doi:10.1093/sleep/zsac317. PMID: 36566505.
70. Banhara FL, Trindade-Suedam IK, Trindade IEK, Sena LMF, Trindade SHK. Robin Sequence and Isolated Cleft Palate are Associated With a High Prevalence of Obstructive Sleep Apnea in School-Aged Children. Cleft Palate Craniofac J. 2026;63(4):828-837. doi:10.1177/10556656251316409. PMID: 39886892.
71. Holder-Espinasse M, Abadie V, Cormier-Daire V, Beyler C, Manach Y, Munnich A, et al. Pierre Robin sequence: a series of 117 consecutive cases. J Pediatr. 2001;139(4):588-90. doi:10.1067/mpd.2001.117784. PMID: 11598609.
72. Izumi K, Konczal LL, Mitchell AL, Jones MC. Underlying genetic diagnosis of Pierre Robin sequence: retrospective chart review at two children's hospitals and a systematic literature review. J Pediatr. 2012;160(4):645-650.e2. doi:10.1016/j.jpeds.2011.09.021. PMID: 22048048.
73. Karempelis P, Hagen M, Morrell N, Roby BB. Associated syndromes in patients with Pierre Robin Sequence. Int J Pediatr Otorhinolaryngol. 2020;131:109842. doi:10.1016/j.ijporl.2019.109842. PMID: 31927149.
74. Firth HV, Hurst JA. Stickler syndrome. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 528–9.
75. Swanson D, Ba'th F, Zavala H, Chinnadurai S, Roby BB. Incidence of mandibular distraction osteogenesis in Stickler Syndrome: Variation due to COL2A1 and COL11A1. Int J Pediatr Otorhinolaryngol. 2021;146:110749. doi:10.1016/j.ijporl.2021.110749. PMID: 34004386.
76. Swanson D, Struyk G, Ba'th F, Chinnadurai S, Roby BB. The Incidence of Velopharyngeal Insufficiency in Stickler Syndrome. Cleft Palate Craniofac J. 2024;61(2):231-234. doi:10.1177/10556656221140675. PMID: 36443936.
77. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Stickler dysplasia. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 97–100.
78. Jones KL, Jones MC, del Campo M. Campomelic dysplasia. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 482–3.
79. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Campomelic dysplasia. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 757–61.
80. Jones KL, Jones MC, del Campo M. Marshall syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 366–7.
81. Jones KL, Jones MC, del Campo M. CHARGE syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 356–7.
82. Jones KL, Jones MC, del Campo M. Oculocerebrofacial syndrome (Kaufman). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 332–3.
83. Jones KL, Jones MC, del Campo M. Boomerang dysplasia. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 472–3.
84. Jones KL, Jones MC, del Campo M. Distichiasis-lymphedema syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 856–7.
85. Jones KL, Jones MC, del Campo M. Nager syndrome (acrofacial dysostosis). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 372–3.
86. Jones KL, Jones MC, del Campo M. Mandibulofacial dysostosis with microcephaly. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 364–5.
87. Jones KL, Jones MC, del Campo M. Diastrophic dysplasia. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 520–1.
88. Jones KL, Jones MC, del Campo M. Yunis-Varón syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 630–1.
89. Jones KL, Jones MC, del Campo M. Sternal malformation–vascular dysplasia spectrum. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 912–3.
90. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Chondrodysplasia with joint dislocations, IMPAD1 gPAPP type. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 266–9.
91. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Yunis-Varón syndrome. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 773–5.
92. Firth HV, Hurst JA. Polyhydramnios. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 774–5.
93. Poets CF, Quante M, Wiechers C. Towards Reaching Consensus in the Diagnosis and Management of Infants With Robin Sequence. Acta Paediatr. 2025;114(8):1755-1757. doi:10.1111/apa.70072. PMID: 40129357.
94. Wiechers C, Goetz S, Kreutzer K, Peters K, Weismann C, LeClair J, et al. Does a clinical assessment of upper airway obstruction in robin sequence correspond to sleep study results?. Sleep Med. 2026;148:109218. doi:10.1016/j.sleep.2026.109218. PMID: 42607588.
95. Zhang L, Fei J, Jia J, Shi X, Qu M, Wang H. Case report of neonate Pierre Robin sequence with severe upper airway obstruction who was rescued by finger guide intubation. BMC Anesthesiol. 2019;19(1):84. doi:10.1186/s12871-019-0754-2. PMID: 31117960. PMC6530166.
96. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Nager syndrome. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 886–8.
97. National Center for Biotechnology Information. ClinVar; VCV000162455 — NM_014305.4(TGDS):c.298G>T (p.Ala100Ser). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/162455/
98. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Oto-spondylo-megaepiphyseal dysplasia. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 103–8.
99. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Catel-Manzke syndrome. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 270–1.
100. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Cerebro-costo-mandibular syndrome. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 790–2.
101. Marques IL, de Sousa TV, Carneiro AF, Peres SP, Barbieri MA, Bettiol H. [Robin sequence: a single treatment protocol]. J Pediatr (Rio J). 2005;81(1):14-22. PMID: 15742081.
102. Wiechers C, Goetz S, Kreutzer K, Weismann C, LeClair J, McGee G, et al. A Randomized Crossover Trial to Evaluate the Effect of Positioning on Obstructive Sleep Apnea in Infants with Robin Sequence. Children (Basel). 2025;12(3). doi:10.3390/children12030389. PMID: 40150671. PMC11941664.
103. de Blacam C, Feeley A, Breugem C, Fauroux B, Javadpour S, Poets C, et al. Systematic Review of Safety and Efficacy of Positional Therapy as First-line Management for Robin Sequence. J Craniofac Surg. 2025;36(8):2932-2937. doi:10.1097/SCS.0000000000011325. PMID: 40162983.
104. Adnot P, Vegas N, Coutier L, Gras Leguen C, Giuseppi A, Aubineau Y, et al. Benefit/risk balance of prone positioning as first line treatment for upper airway obstruction in Robin sequence. Eur J Pediatr. 2025;185(1):21. doi:10.1007/s00431-025-06678-5. PMID: 41400683. PMC12708752.
105. Équipe du CRMR SPRATON Paris. Précautions pour le couchage ventral d'un bébé avec séquence de Pierre Robin [parent information sheet]. Paris: Centre de Référence SPRATON, Hôpital Necker-Enfants malades; 2025 Jun. Available from: https://www.tete-cou.fr/medias/fichiers/fichier_spraton_necker_fiche_couchage_ventral_2025.pdf
106. Thean DC, Jindal S, Resnick CM. Barriers to Adoption of the Pre-Epiglottic Baton Plate (Tübingen Palatal Plate) for Infants With Robin Sequence. Cleft Palate Craniofac J. 2026;63(9):2834-2840. doi:10.1177/10556656251393877. PMID: 41204099.
107. Case HF, Layuno Matos M, Tang K, Boven L, Brodie K, Barbour M, et al. Descriptive Summary and Experience Introducing the Orthodontic Airway Plate to Treat Upper Airway Obstruction for Infants With Robin Sequence. Cleft Palate Craniofac J. 2026;63(9):3059-3068. doi:10.1177/10556656251394070. PMID: 41329835. PMC13162521.
108. Giuseppi A, Vegas N, Thierry B, Soupre V, Donatien A, Adnot P, et al. Implementing the Palatal Plate for Robin Sequence Beyond the Originating Center: How I Do It. Laryngoscope. 2026;136(6):2801-2804. doi:10.1002/lary.70411. PMID: 41631678. PMC13274758.
109. Selvaraju K, Tan CS, Hariri F, Saniasiaya J. Outcome of orthodontic airway plate in improving airway among children with Pierre Robin Sequence: a systematic review. J Laryngol Otol. 2026;140(3):310-317. doi:10.1017/S002221512510412X. PMID: 41376129.
110. Myers S, Beiriger JW, Bruce MK, Matinrazm S, Dvoracek L, Raghuram A, et al. True Incidence of Marginal Mandibular Nerve Palsy following Neonatal Mandibular Distraction Osteogenesis. Plast Reconstr Surg. 2025;155(2):317-324. doi:10.1097/PRS.0000000000011409. PMID: 38546544.
111. Lakhsmi Putri I, Islami K, Fitri Aisyah I, Pramanasari R, Kencono Wungu CD. Mandibular Distraction Osteogenesis vs. Tracheostomy in the Management of Pierre Robin Sequence: A Systematic Review and Meta-Analysis. Cleft Palate Craniofac J. 2026;63(4):700-710. doi:10.1177/10556656241311549. PMID: 39782888.
112. Resnick CM, de Gijsel J, Jindal S, Quante M, Poets C, Katz E, et al. Comparative Effectiveness of Preepiglottic Baton Plates and Mandibular Distraction in Infants with Robin Sequence. Plast Reconstr Surg. 2025;156(2):267e-278e. doi:10.1097/PRS.0000000000011887. PMID: 39560157. PMC12288888.
113. Dominguez RW, Velazquez AE, Cinclair T, Wang CS, Kane AA, Sanati-Mehrizy P. Mandibular Distraction Osteogenesis for Pierre Robin Sequence in Early Infancy: A Systematic Review. Plast Reconstr Surg. 2026;:doi: 10.1097/PRS.0000000000012852. doi:10.1097/PRS.0000000000012852. PMID: 41586544.
114. Gómez OJ, Barón OI, Peñarredonda ML. Pierre Robin Sequence: An Evidence-Based Treatment Proposal. J Craniofac Surg. 2018;29(2):332-338. doi:10.1097/SCS.0000000000004178. PMID: 29215441.
115. Resnick CM, LeVine J, Calabrese CE, Padwa BL, Hansen A, Katwa U. Early Management of Infants With Robin Sequence: An International Survey and Algorithm. J Oral Maxillofac Surg. 2019;77(1):136-156. doi:10.1016/j.joms.2018.05.031. PMID: 30599883.
116. van de Velde S, Smit MM, Logjes RJH, Martin E, Haasnoot M, Mink van der Molen AB, et al. Postoperative respiratory difficulties following primary cleft palate repair in infants with Robin sequence versus isolated cleft palate: A retrospective study. J Craniomaxillofac Surg. 2024;52(12):1441-1448. doi:10.1016/j.jcms.2024.08.021. PMID: 39256140.
117. Kiefer H, Poets CF, Wiechers C, Neukamm AC, Naros A, Finke H, et al. Health-related quality of life in children with Robin sequence following Tübingen palatal plate treatment in infancy. Arch Dis Child. 2026;111(10):897-903. doi:10.1136/archdischild-2026-330540. PMID: 42425756. PMC13638337.
118. Rosen R, Vandenplas Y, Singendonk M, Cabana M, DiLorenzo C, Gottrand F, et al. Pediatric Gastroesophageal Reflux Clinical Practice Guidelines: Joint Recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (NASPGHAN) and the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN). J Pediatr Gastroenterol Nutr. 2018;66(3):516-554. doi:10.1097/MPG.0000000000001889. PMID: 29470322. PMCID: PMC5958910.
119. AstraZeneca Pharmaceuticals LP. NEXIUM (esomeprazole magnesium) delayed-release capsules and for delayed-release oral suspension. Prescribing information (SPL of 2025-12-19; set id f4853677-1622-4037-688b-fdf533a11d96) [Internet]. U.S. Food and Drug Administration / openFDA [cited 2026 Nov 6]. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=f4853677-1622-4037-688b-fdf533a11d96
120. U.S. National Library of Medicine. ClinicalTrials.gov [Internet]. Search of records with the term "Robin" (16 records, including NCT06178562, NCT02432638, NCT02266043, NCT03423017, NCT07604818 and NCT07257276). Accessed 2026 Oct 10. Available from: https://clinicaltrials.gov/
121. Lee MS, Guelmann M, Sheller B. Orthodontic Airway Plate Therapy for Infants with Robin Sequence: Clinical Principles, Current Evidence, and Evolving Innovations. Neonatology. 2026;:1-11. doi:10.1159/000552259. PMID: 42033735. PMC13252920.
122. Kapur A, Tamchos R, Mathew JL, Kumar J, Pavithra K, Zalmi K, et al. Clinical experience and functional outcomes of infants with Pierre Robin sequence treated with the Tübingen palatal plate: a retrospective case series from a multidisciplinary tertiary care setting. Eur Arch Paediatr Dent. 2026;:doi: 10.1007/s40368-026-01263-2. doi:10.1007/s40368-026-01263-2. PMID: 42616257.
123. Tabchi Y, Bentahar O. Mandibular distraction osteogenesis in Robin sequence: identifying the optimal timing: A systematic review. Int J Pediatr Otorhinolaryngol. 2026;207:112908. doi:10.1016/j.ijporl.2026.112908. PMID: 42320445.
124. Goel D, Wilson A, Rao S, Gillett D, Hayley Herbert H, O'Donnell A. Robin Sequence: From Dilemmas to Developing an Adaptable Standardized Stepwise Approach. Acta Paediatr. 2025;114(8):1760-1777. doi:10.1111/apa.70028. PMID: 39945272. PMC12258098.
125. Paes EC, van Nunen DP, Basart H, Don Griot JP, van Hagen JM, van der Horst CM, et al. Birth prevalence of Robin sequence in the Netherlands from 2000-2010: a retrospective population-based study in a large Dutch cohort and review of the literature. Am J Med Genet A. 2015;167A(9):1972-82. doi:10.1002/ajmg.a.37150. PMID: 25994858.
126. Vatlach S, Maas C, Poets CF. Birth prevalence and initial treatment of Robin sequence in Germany: a prospective epidemiologic study. Orphanet J Rare Dis. 2014;9:9. doi:10.1186/1750-1172-9-9. PMID: 24433508. PMC3899445.
127. Borg D, Chircop KL, Attard Z, Parnis J. Epidemiology and Incidence of Pierre Robin Sequence in the Maltese Islands: A Population-Based Study. J Craniofac Surg. 2025;36(8):2943-2946. doi:10.1097/SCS.0000000000011277. PMID: 40172971.