Endocrine-Cerebro-Osteodysplasia Syndrome and ICK-Related Ciliopathies

Autosomal recessive skeletal ciliopathy, lethal in the perinatal period, caused by biallelic variants in the catalytic domain of the CILK1 kinase (formerly ICK), with hydrocephalus, holoprosencephaly, median cleft lip and palate, micromelia, polydactyly, pituitary and adrenal hypoplasia and genital anomalies; it includes the short-rib polydactyly phenotype of the same gene and, as allelic forms, cranioectodermal dysplasia type 6 and susceptibility to juvenile myoclonic epilepsy 10.

Literature and regulatory information updated: 31 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Definition and place among the ciliopathies

Endocrine-cerebro-osteodysplasia (ECO) syndrome is an autosomal recessive disease, lethal in the neonatal period, with multiple anomalies involving three systems: osteodysplasia (micromelia, bowing of the long bones, postaxial polydactyly), brain anomalies (hydrocephalus, holoprosencephaly, absence or hypoplasia of the corpus callosum, cortical malformations) and hypoplasia of the endocrine glands, particularly the pituitary and adrenals, together with median cleft lip and palate and ambiguous genitalia [1]. Orphanet defines it as a malformation syndrome with anomalies of the endocrine, cerebral and skeletal systems resulting in neonatal mortality, with onset in the neonatal period or in infancy [2]. ECO belongs to the ciliopathies, the group of genetic diseases caused by altered ciliary function, and in particular to the ciliopathies with major skeletal involvement, in which primary cilia, single sensory organelles on each cell, coordinate developmental signalling pathways such as sonic hedgehog (SHH) [23, 24]. Primary cilia act as the cell's antennae and take part in breaking embryonic symmetry, which explains why ciliopathies affect the development of many organs at the same time [23].

1.2. Founding description

In 2009, Lahiry and colleagues, at the University of Western Ontario, described six affected fetuses and newborns, born between 2002 and 2006 in two sibships of an Old Order Amish family with two consanguineous marriages, with a set of endocrine, cerebral and skeletal anomalies that they considered a new and distinct syndrome, which they named ECO [11]. One newborn died on the third day of life, another during delivery, and the remaining pregnancies were terminated between 23 and 29 weeks for severe malformations identified on prenatal ultrasound [1, 11]. By autozygosity mapping in five affected individuals and three of the four parents, the authors localised the gene to 6p12.2–p12, in a region of about 1.8 Mb, and identified a homozygous missense variant in the ICK gene (intestinal cell kinase), p.(Arg272Gln), with complete linkage to the disease (two-point LOD score, logarithm of the odds of linkage, of 3.61 at zero recombination fraction) [11]. The authors noted similarities with Majewski-type short rib-polydactyly syndrome and with hydrolethalus syndrome, but the loci of these did not overlap with that of ICK [11].

1.3. Later milestones

In 2016, Oud and colleagues described the second family worldwide: a Turkish fetus of consanguineous parents, initially diagnosed as short-rib thoracic dysplasia (SRTD) type II, with a different homozygous variant in the same gene, p.(Gly120Cys), and showed reduced ciliogenesis in fibroblasts from an Amish patient, confirming that ECO is a severe ciliopathy [13]. In the same year, Paige Taylor and colleagues identified, in a cohort of the International Skeletal Dysplasia Registry at the University of California, Los Angeles, a newborn with a short rib-polydactyly syndrome (SRPS) phenotype and the homozygous variant p.(Glu80Lys), which abolishes kinase activity, and concluded that ICK variants produce an allelic spectrum ranging from ECO to SRPS [15]. In 2018, Bailey and colleagues associated rare heterozygous variants in the same gene with juvenile myoclonic epilepsy (JME), in a multigenerational family from Belize and in a cohort of 310 probands [12]. In 2025, Sezer and colleagues described a frameshift variant in the non-catalytic C-terminal domain in five children from two Iraqi Turkmen families with cranioectodermal dysplasia, who survived childhood [14, 25].

1.4. Nomenclature and reclassifications

The kinase was cloned three times: from a brain complementary DNA library in 1999 (KIAA0936), from colonic mucosa in 2000, where it is expressed in the intestinal crypts and from which the name “intestinal cell kinase” derives, and from laryngeal carcinoma cells in 2002, under the name MRK (MAK-related kinase) [5]. Because the protein is ubiquitously expressed and acts in the primary cilium, the symbol approved by the HGNC (HUGO Gene Nomenclature Committee) became CILK1 (ciliogenesis-associated kinase 1) [5, 26]. The clinical literature up to 2018 uses ICK, and later texts use CILK1; this article uses CILK1 for the gene and keeps ICK in historical citations. OMIM keeps phenotype entry 612651 for ECO, which also includes the patient with an SRPS phenotype, and created in 2018 entry 617924 for susceptibility to EJM10 and in 2025 entry 621337 for cranioectodermal dysplasia type 6 (CED6) [1, 7, 14]. The 2023 Nosology of Genetic Skeletal Disorders lists “endocrine-cerebro-osteodysplasia (ECO), CILK1-related” in group 10, between the SRPS and the SRTD [3]. In November 2025, ClinGen grouped the two recessive forms (ECO and CED6) into a single entity, CILK1-related ciliopathy, and split off the dominant form (EJM10), on the basis of differences in inheritance, molecular mechanism and phenotype [4].

1.5. Entities within scope

This article covers ECO as the main entity and describes the allelic spectrum of CILK1 uniformly: classic ECO due to missense variants in the kinase domain (p.(Arg272Gln) and p.(Gly120Cys)), the SRPS phenotype due to complete loss of catalytic activity (p.(Glu80Lys)), CED6 due to a distal truncating variant and susceptibility to EJM10 due to heterozygous variants [5]. CED6 (cranioectodermal dysplasia or Sensenbrenner syndrome) and the genetic epilepsies in general enter the differential diagnosis. As ECO is diagnosed mainly in the context of a fetal or neonatal skeletal dysplasia with a narrow thorax and polydactyly, the differential diagnosis with the SRPS, the SRTD, hydrolethalus syndrome and the oral-facial-digital syndromes takes up part of the article, alongside Joubert syndrome, Meckel-Gruber syndrome, nephronophthisis and the related renal ciliopathies, Bardet-Biedl syndrome and the polydactylies of genetic cause.

2. Molecular Genetics and Pathophysiology

2.1. The gene and the protein

CILK1 is located at 6p12.1 (GRCh38 chr6:53,001,303–53,061,824; GRCh37 chr6:52,866,101–52,926,622) and the MANE Select reference transcript is NM_014920.5 (ENST00000676107.1), which encodes the protein NP_055735.1 [4, 27]. The protein has 632 amino acids and a calculated mass of 71.4 kDa; the N-terminal catalytic domain has 12 subdomains and includes in subdomain VIII a dual-phosphorylation TDY motif, similar to that of the mitogen-activated protein kinases (MAPK); this is followed by two putative nuclear localisation signals and a proline-rich SH3-binding domain [5]. The non-catalytic C-terminal region is long and intrinsically disordered, and is required for substrate binding and for targeting to the cilium [28]. The catalytic domain shares 38–40% identity with that of the cyclin-dependent kinases (CDK) and the MAPK, and arginine 272 is conserved in the 84 known orthologues and in the CMGC group of kinases (CDK, MAPK, GSK3, CLK) [11]. There is an ICKb isoform, 337 amino acids shorter and with an alternative C-terminus, mainly cytoplasmic [11]. Expression is ubiquitous; the transcript of about 6 kb is most abundant in placenta, thymus and small intestine, or in brain, spinal cord, testis and ovary, depending on the method, and in the brain the corpus callosum stands out [5]. In the developing mouse brain, the kinase is expressed in the ependyma of the lateral ventricles, the choroid plexus, the hippocampus, the neocortex and the Purkinje cells, and in growth cartilage mainly in proliferative and prehypertrophic chondrocytes, with cytoplasmic and nuclear distribution [5, 15].

2.2. Function in the primary cilium

CILK1 belongs to the RCK (v-ros cross-hybridizing kinase) family of serine/threonine kinases, together with MAK (male germ cell-associated kinase) and MOK, and its homologues in Caenorhabditis elegans (dyf-5), Chlamydomonas reinhardtii (LF4p) and Leishmania mexicana (LmxMPK9) are negative regulators of cilium and flagellum length [15]. Activation depends on autophosphorylation of tyrosine 159, which gives basal activity, and on phosphorylation of threonine 157 by CCRK (cell cycle-related kinase, also CDK20), which gives full activation; phosphorylation of tyrosine 15 partially inactivates it, and the phosphatase PP5 deactivates it [11, 15, 28]. The KATNIP protein binds the disordered C-terminal region at the basal body and acts as a regulatory subunit, increasing CILK1 levels and TDY phosphorylation, phosphorylation of its substrates and suppression of ciliary length [29]. KATNIP variants that cause Joubert syndrome bind CILK1 but do not activate it, mechanistically linking CILK1 to this ciliopathy [30]. Candidate substrates include the kinesin KIF3A, the motor of anterograde intraflagellar transport (IFT), the antiapoptotic protein Scythe (BAT3) and the mTORC1 complex [11, 28]. Lack of CILK1 accelerates anterograde IFT and lengthens cilia [28]. In the mouse retina, Mak and Ick cooperate as IFT regulators at the ciliary tip of photoreceptors, activated by Ccrk, and loss of both kinases destroys the axonemes and causes retinal degeneration [31].

2.3. Mechanism of the ECO and SRPS variants

The p.(Arg272Gln) variant of the Amish patients disrupts the ionic bond between arginine 272 and glutamate 189, which becomes exposed on the surface; the Rosetta and Eris programs predict a loss of stability of 5.1 and 3.4 kcal/mol, respectively [11]. The mutant protein localises to the nucleus in 9.78% of transfected cells, versus 71.9% for the wild type, and does not phosphorylate myelin basic protein in vitro [11]. Measured by autophosphorylation of tyrosine 159, the variant retains about one third of basal activity [15]. The p.(Gly120Cys) variant of the Turkish fetus, also in the catalytic domain, retains nuclear localisation, but both variants accumulate at a bulging ciliary tip in about 75% of ciliated cells, whereas the wild-type protein is distributed along the axoneme with enrichment at the base in 92%; in fibroblasts from a patient with p.(Arg272Gln) ciliogenesis is reduced, with cilia of normal length [13]. The p.(Glu80Lys) variant replaces a negatively charged residue with a positively charged one at the ATP-binding interface, abolishes kinase activity and autophosphorylation and shifts the protein to the ciliary tip, and the patient's fibroblasts have thin, long and twisted cilia [15]. In these fibroblasts, GLI3 processing is altered, with an increase in the full-length form relative to the repressor form and reduced expression of GLI1 and PTCH1; as the balance between the two forms of GLI3 determines digit specification, this imbalance is the proposed explanation for the polydactyly [15]. Basal ERK is increased in the fibroblasts, but its inhibition does not shorten the cilia, and the growth plates of Ick−/− mice have, on the contrary, less phosphorylated ERK [15]. The authors interpret the difference between ECO and SRPS as dependent on the degree of residual catalytic activity: about one third with p.(Arg272Gln), almost none with p.(Glu80Lys) [15]. Lahiry and colleagues further suggested that several malformations (cleft lip and palate, syndactyly, persistent eyelid fusion, unfused urogenital folds) reflect an apoptosis defect, given the phosphorylation of Scythe by the kinase [11].

2.4. Animal models

Ick−/− mice have a disorganised growth plate, with a short proliferative zone, poor hypertrophic columns, small chondrocytes and less extracellular matrix, fewer trabeculae and mid-diaphyseal bowing of the long bones, with histological similarities to thanatophoric dysplasia [5, 15]. Mice homozygous for the p.(Arg272Gln) variant (knock-in) die at birth from respiratory failure, with reduced lung branching, airspace deficiency, abnormal mesenchymal differentiation, elongated cilia and altered hedgehog signalling and autophagy [32]. The same model reproduces the short limbs and polydactyly, and reveals spinal deformity with a defective intervertebral disc, reduced mineralisation of the spine, ribs and long bones and fewer hypertrophic chondrocytes expressing type X collagen [33]. The cleft palate of the mutant mice results from reduced proliferation of the palatal shelves, with normal adhesion and fusion [34]. In targeted deletion models, lack of Cilk1 in the renal collecting ducts causes polycystic kidney disease, lack in the intestinal mesenchyme prevents ciliary maintenance and villus morphogenesis, and lack in the teeth produces diastemal supernumerary teeth dependent on the level of hedgehog signalling [35–37].

2.5. Inheritance and allelic spectrum

ECO and CED6 are autosomal recessive; in the published families the parents were heterozygous and unaffected [11, 13, 14]. EJM10 is catalogued as autosomal dominant with incomplete penetrance, in the form of susceptibility [7]. The same gene thus has two inheritance patterns, linked to different types of variants and distinct phenotypes, and ClinGen assesses them as separate entities [4]. The following table summarises the entities.

2.6. Variants and nomenclature

The HGVS (Human Genome Variation Society) recommends describing variants relative to a versioned reference transcript, preferably the MANE Select. The CILK1 literature has used different transcripts, and the variant of the Amish patients appears in the founding article and in OMIM as “c.1305G>A”, in a numbering counted from the start of the complementary DNA (GI 156671211) and not from the initiation codon [5, 11]. In VariantValidator, NM_016513.5:c.1305G>A corresponds to a synonymous variant, p.(Leu435=), so that notation should not be used; the correct description is NM_014920.5:c.815G>A p.(Arg272Gln), already used by Oud and colleagues in 2016 [13, 27]. The following table brings together the variants cited, normalised in VariantValidator (GRCh38) and with their ClinVar classification at the date of consultation (31 October 2026).

ClinVar also contains variants submitted without an identified associated publication and classified by a single laboratory as likely pathogenic for ECO, NM_014920.5:c.1024C>T p.(Gln342Ter) (VCV005002900) and NM_014920.5:c.-172-1G>A (VCV004526597), and NM_014920.5:c.271A>T p.(Lys91Ter) as likely pathogenic for a “CILK1-related disorder” (VCV004995130), which indicates the existence of additional cases whose phenotype has not been described in the literature [38]. The large 6p12 deletions that include the gene, recorded as pathogenic, are not attributable to CILK1 alone [38].

2.7. Founder variants

The c.815G>A variant was found in 5 of 257 healthy Old Order Amish controls, all heterozygous and from outside the affected family, giving an allele frequency of 0.97% and a carrier frequency of about 1 in 51; the authors estimated that about 1 in 10,000 Amish would be homozygous [11]. The variant was not found in 2855 individuals from six non-Amish ethnic groups, which makes it a variant private to this community [11]. The p.(Gly120Cys) variant was not present in 72 in-house Turkish exomes, in 1020 Turkish exomes from the national TUBITAK project or in the ExAC database [13]. The CED6 variant lies on a shared haplotype of about 19 Mb, consistent with a founder effect in the Iraqi Turkmen population [5, 25].

2.8. Genotype-phenotype correlation

The ECO and SRPS variants lie in the kinase domain and reduce or abolish catalytic activity, whereas the CED6 variant lies in the non-catalytic C-terminal domain, escapes nonsense-mediated messenger RNA decay because it is in the last exon, and is associated with survival [5, 25]. In CED6, fibroblast cilia are about twice as long and IFT components (IFT70, IFT88) accumulate at the tip, indicating a retrograde transport defect [5]. With only three ECO families published, the correlation between the degree of residual activity and the predominance of holoprosencephaly (Amish) or short-rib dysplasia (p.(Glu80Lys)) is a hypothesis of the authors, not a demonstrated rule [15].

2.9. The association with juvenile myoclonic epilepsy

Bailey and colleagues identified, in a 37-member family from Belize of European and Amerindian ancestry with 12 living affected members over four generations, linkage to 6p12.2 (maximum LOD 3.35) and the heterozygous variant p.(Lys305Thr), which cosegregated with the 12 affected members and was also present in 3 unaffected relatives [12]. On screening the gene in 310 probands with JME from the GENESS consortium (Genetic Epilepsies Studies, with centres in the United States, Mexico, Honduras, Brazil and Japan), 24 had variants (8%), 21 variants absent or extremely rare in gnomAD were found in 22 patients (7%), 13 of the 24 variants were significant in case-control analysis with an odds ratio above 5, and only four (p.(Lys305Thr), p.(Lys220Glu), p.(Ala615Thr) and p.(Arg632Ter)) were classified as pathogenic by ACMG/AMP criteria (American College of Medical Genetics and Genomics / Association for Molecular Pathology) [7, 12]. There was cosegregation in 9 families and lack of segregation in 3 [12]. In electroporation experiments in the embryonic mouse neocortex, the variants altered radial migration, mitosis and cell cycle exit and increased apoptosis (except p.(Ala615Thr)), and Ick+/− mice had more tonic-clonic seizures under isoflurane (P=0.02) and more diffuse polyspikes (P=0.047), which led the authors to propose haploinsufficiency [12]. This interpretation has been challenged: in an independent study with prolonged electrocorticographic recordings, neither Cilk1+/− mice nor mice heterozygous for p.(Arg272Gln) had spontaneous seizures or seizures provoked by recovery from isoflurane, and heterozygous cells have normal cilia [39]. Cellular studies show that the JME variants lose the ability to limit ciliary length and gain the ability to promote ciliogenesis, and that the murine equivalent of p.(Ala615Thr) increases ciliation and hedgehog signalling with a single allele and impairs regulation by KATNIP, suggesting a mechanism different from simple haploinsufficiency [28, 40]. The population indices of the gene (pLI 0 and LOEUF 0.78 in gnomAD v4) do not indicate intolerance to loss of function [4]. In ClinVar, the four EJM10 variants are listed as of uncertain significance or as risk factors, and ClinGen split off EJM10 for assessment by the epilepsy expert panel [4, 38].

3. Clinical Spectrum and Manifestations by System

3.1. Data basis

The clinical description of ECO rests on three families: six Old Order Amish fetuses and newborns born between 2002 and 2006, described individually by Lahiry and colleagues, to whom Oud and colleagues added a seventh affected member of the same family in an updated table; a Turkish fetus terminated at 33 weeks; and a newborn with an SRPS phenotype who died hours after delivery [11, 13, 15]. The denominators are therefore very small, several features were only sought in cases with autopsy, and the percentages should not be read as population frequencies.

The table summarises the data from [11, 13, 15].

3.2. Central nervous system

The neuropathological changes described in the Amish patients with autopsy are extensive: holoprosencephaly (failure of separation of diencephalic elements, a monoventricular semilobar form with fusion of the basal ganglia and thalami in one case, diencephalic agenesis in another), agenesis or hypoplasia of the corpus callosum, absent septum pellucidum, dysmorphic or stenotic cerebral aqueduct, absent olfactory bulbs, malformation of the frontal and occipital cortex with focal polymicrogyria or rudimentary sulcation, brainstem malformation, a small or haemorrhagic cerebellum or one with dysmorphic peduncles, hippocampal agenesis, leptomeningeal glioneuronal heterotopia and spinal cord malformation [11]. The hydrocephalus presented as ventricular dilatation with absence of the third ventricle; the authors proposed that overgrowth of the medial diencephalic structures obliterates the third ventricle and causes secondary hydrocephalus, with rupture or absence of the septum pellucidum [11]. In one fetus, the hydrocephalus required drainage of 350 ml of cerebrospinal fluid from the ventricles to allow vaginal delivery [11]. Head circumference was above the 97th percentile in two of the Amish and in the Turkish fetus [11, 13].

3.3. Face and oral cavity

The face is characterised by midface hypoplasia and a broad flat nasal root (6/6), micrognathia (6/6), small deep-set eyes, dysplastic low-set ears, excess submental skin and a laterally deficient lower lip [11]. Median cleft palate and median or bilateral cleft lip are almost constant, the premaxilla may be absent, and there are reports of hypoplasia or absence of the epiglottis and larynx and of two supernumerary teeth at 29 weeks [11]. The Turkish fetus had a high broad forehead, hypertelorism, a very small nose with anteverted nares, a median pseudocleft of the upper lip, soft palate cleft, natal teeth, labiogingival fusion, an irregular alveolar ridge, tongue fused to the lower palate, multiple frenula and low-set, posteriorly rotated ears, which brought the picture close to the oral-facial-digital syndromes [13]. The newborn with p.(Glu80Lys) had an anterior fontanelle of 7×7 cm with widely separated sutures, frontal bossing, a short upturned nose with tented alae, microstomia and ears without an antihelix [15].

3.4. Skeleton

All patients have micromelia, with bowing of the radius and ulna, ulnar deviation of the hands, brachydactyly, syndactyly, abducted hips and a gap between the first and second toes; bowing of the legs was seen in 2 of 6 Amish [11]. In the only Amish patient whose radiographs were analysed by the International Skeletal Dysplasia Registry, the diaphyses were short, the ulnae short and bowed, the tibiae short and ovoid, the femur short and ovoid, the ilium abnormal and the acetabular roof hypoplastic [11]. The Turkish fetus had very short tubular bones, torpedo-shaped femora, hypoplastic ovoid tibiae with very short fibulae, premature epiphyseal ossification at the knees, short ribs, slightly deficient vertebral ossification and small ilia of normal contour, a combination similar to SRTD type II (Majewski) [13]. The newborn with p.(Glu80Lys) had a long narrow thorax with short horizontal ribs, relatively long clavicles, short humeri and femora, bowed radii, ulnae, tibiae and fibulae, smooth metaphyses and a long narrow ilium with a smooth acetabular roof and high ischia [15]. Polydactyly is postaxial in the Amish and in the patient with p.(Glu80Lys), and combined (postaxial in the hands, preaxial in the feet) in the Turkish fetus [11, 13, 15].

3.5. Endocrine system and genitalia

The pituitary was absent or unidentifiable in the two Amish patients with autopsy who described it, and the adrenals were hypoplastic in two and absent in one [11]. In one fetus, integrated prenatal screening was positive for trisomy 21, with unconjugated oestriol of 0.18 MoM (multiples of the median) and alpha-fetoprotein of 0.48 MoM, and the authors suggested that adrenal hypoplasia explained the low oestriol [11]. In the 46,XY fetuses, differentiation of the external genitalia was abnormal to varying degrees: microphallus with hypoplastic scrotum, bilateral cryptorchidism and, in one case, sex reversal with hypoplastic labia majora; the Turkish fetus also had sex reversal, and in the patient with p.(Glu80Lys) the external genitalia were absent [11, 13, 15]. One of the Amish had squamous metaplasia of the urethra and the periurethral prostatic glands [11]. No postnatal hormone measurements have been published.

3.6. Other organs

The kidneys of the Amish patients showed cystically dilated tubules in the medulla and cortex, and the liver had persistent circular ductal plates, a sign of ductal plate malformation [11]. The Turkish fetus had large hyperechogenic kidneys at 19 weeks, whereas the patient with p.(Glu80Lys) had no renal anomalies [13, 15]. A tubular or small stomach and pulmonary hypoplasia were described and, in the Amish, absence of heart disease; the Turkish fetus had a ventricular septal defect [11, 13]. Weight was above the 90th percentile in four of the six measurements in the Amish, which the authors attributed in part to the excess fluid of severe hydrocephalus [11].

3.7. Pregnancy and the perinatal period

In the Amish pregnancies in which ultrasound was performed, the multiple anomalies (ventriculomegaly, cleft lip, short limbs) were detected in the second trimester; there was polyhydramnios in one pregnancy and spontaneous labour in two, at 33–34 weeks [11]. In the Turkish fetus, the 16-week ultrasound showed cystic hygroma, scalp oedema, ascites, bilateral ventriculomegaly, very short tubular bones, polydactyly and short ribs, and the 19-week scan large hyperechogenic kidneys and a ventricular septal defect [13]. One of the Amish newborns had no spontaneous breathing, was intubated for one day, extubated and taken home for palliative care, and died on the third day [11]. The newborn with p.(Glu80Lys) was born at 38 weeks weighing 4050 g, 44 cm long and with Apgar scores of 2, 6 and 7, was intubated for respiratory failure and died within a few hours [15].

3.8. Allelic form with survival: cranioectodermal dysplasia 6

The five children with CED6 survived childhood; four were alive at 5 months and at 3, 7 and 10 years, and one died of chronic liver disease at 9 years [14]. All had disproportionate short stature with brachydactyly of the fingers and toes and joint laxity, ectodermal dysplasia (sparse hair, eyebrows and eyelashes, dental anomalies, lax skin), renal involvement (hyperechogenic kidneys, pyelectasis and, in some, end-stage chronic kidney disease) and liver disease (from raised transaminases to cirrhosis); four had relative macrocephaly with external hydrocephalus [14]. In the six-generation pedigree of one of the families there were 11 further relatives with unconfirmed suspicion, all deceased, most between 1 and 13 years of age [14]. There was no epilepsy in either family [14].

3.9. Dominant allelic form: juvenile myoclonic epilepsy 10

JME is the most frequent genetic generalised epilepsy and the most common cause of seizures in adolescents and young adults; it is characterised by myoclonic jerks on awakening, especially after sleep deprivation or alcohol, and by generalised polyspikes on the electroencephalogram (EEG) [12]. The International League Against Epilepsy (ILAE) includes it, in 2022, among the four idiopathic generalised epilepsy syndromes [41]. In the Belize family with p.(Lys305Thr), of the 12 affected members, 3 had JME, 2 only myoclonic-tonic-clonic seizures, 2 febrile seizures with childhood absences or neonatal myoclonus, 1 only febrile seizures and 4 were asymptomatic with polyspikes on EEG [12]. Among the 34 affected non-proband relatives in all the families with variants, 5 (15%) had JME, 10 (29%) myoclonic-tonic-clonic seizures, 4 (12%) pyknoleptic absences with or without these seizures, 4 (12%) febrile seizures alone or with absences or myoclonus, and 11 (32%) were asymptomatic with polyspikes or focal spikes on EEG [7, 12]. The EEG typically shows 3.5–5 Hz polyspike-wave complexes, age at onset varies between the first years of life and adolescence, and no patient had photosensitivity [7]. The sources consulted do not describe epilepsy in the heterozygous parents of children with ECO, but nor do they state that it was looked for.

3.10. Natural history and treated disease

All published cases of ECO ended in fetal death, death during delivery or in the first hours or days of life, or in termination of pregnancy for medical reasons, and there is no description of treated disease or prolonged survival [11, 13, 15]. Lethality results from the combination of brain malformations, pulmonary hypoplasia and, in the short-rib phenotypes, thoracic restriction, as in the SRPS, in which death from respiratory failure secondary to pulmonary hypoplasia occurs in the first days in all infants [24, 32]. The absence of survivors in three families does not show that hypomorphic CILK1 variants are always lethal, as CED6 demonstrates [14].

4. Clinical Recognition and Red Flags

4.1. Prenatal period

ECO should be considered when second-trimester ultrasound shows the association of ventriculomegaly or hydrocephalus, median cleft lip or palate, short bowed long bones and polydactyly, with or without short ribs, hyperechogenic kidneys, cystic hygroma or hydrops [11, 13]. The combination of midline brain anomalies (holoprosencephaly, corpus callosum agenesis) with short-rib skeletal dysplasia and polydactyly is what distinguishes ECO from most SRTD, in which major brain malformations are less frequent [11, 17]. Biochemical screening with very low oestriol may reflect adrenal hypoplasia, but it was observed in a single fetus and does not serve as a marker [11]. Parental consanguinity, membership of a founder community (Old Order Amish) and a history of a previous fetus with skeletal dysplasia and brain malformations increase the prior probability [11, 13]. When the long bones are below the 5th percentile or 2 standard deviations below the mean before 24 weeks, referral to a centre experienced in skeletal dysplasias and genetic counselling is recommended [42].

4.2. Newborn

In the newborn, the signs pointing to a skeletal ciliopathy are a narrow bell-shaped thorax, short ribs, short limbs, polydactyly, anomalies of the mouth, tongue and frenula and small or ambiguous genitalia [23]. The association with hydrocephalus, holoprosencephaly, eyelid fusion, median cleft and adrenal or pituitary hypoplasia points to ECO [1, 11]. A history of stillbirths or neonatal deaths in the family and consanguinity form part of the history-taking for any suspected ciliopathy [23].

4.3. Emergency signs

Respiratory failure at birth, due to pulmonary hypoplasia and thoracic restriction, is the dominant emergency and the usual cause of death [15, 24]. Severe hydrocephalus may determine the mode of delivery [11]. Hypoplasia or absence of the adrenals and pituitary, documented at autopsy, implies a risk of adrenal insufficiency and hypopituitarism in a newborn who survives the first hours; none of the sources describes hormonal assessment or replacement in living patients, so their assessment follows the general principles of neonatology [11].

4.4. Juvenile myoclonic epilepsy

In JME, suspicion of a monogenic cause with family implications is rare, and the association with CILK1 is not established for clinical use [4, 38]. The identification of a rare heterozygous CILK1 variant in a patient with JME on an epilepsy panel does not in itself justify testing relatives for predictive purposes [38].

4.5. Newborn screening

There is no newborn screening for ECO or for the skeletal ciliopathies.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no formal clinical criteria for ECO. The diagnosis is molecular: two pathogenic or likely pathogenic CILK1 variants in trans (homozygous in the three published families) in a fetus or newborn with a short-limbed skeletal dysplasia and polydactyly associated with midline brain malformations, median cleft or endocrine hypoplasia [1, 4]. The validity of the gene-disease association for the recessive form is “definitive” according to ClinGen, which allows biallelic CILK1 variants to be interpreted in a diagnostic context [4].

5.2. Prenatal assessment of skeletal dysplasia

The recommendations of the International Skeletal Dysplasia Society published in Genetics in Medicine in 2009 state that, when the long bones are at or below the 5th percentile, or more than 3 standard deviations below the mean, the fetus should be assessed at a centre experienced in skeletal dysplasias [42]. The skull (biparietal diameter and head circumference), facial profile, mandible, clavicle, scapula, thoracic circumference, vertebral bodies, all long bones and the hands and feet should be measured and compared with normative values; suspicion is strong when the long bones are more than 3 standard deviations below the mean and head circumference is above the 75th percentile [42]. Lethality is estimated from the ratio of thoracic to abdominal circumference, in which a value below 0.6 strongly suggests a perinatal lethal disorder, and from the ratio of femur length to abdominal circumference, in which a value of 0.16 has the same meaning, with known exceptions [42]. The fetal skeleton is visible on two-dimensional ultrasound from 14 weeks [42]. A systematic ultrasound approach remains necessary in the exome era, because variant interpretation depends on precise phenotyping [43]. The 2009 authors advise against in utero radiography, because of insufficient resolution [42].

5.3. Postnatal and post-mortem assessment

Every fetus or newborn with suspected skeletal dysplasia should have clinical and radiological confirmation after delivery: anteroposterior radiographs of the appendicular skeleton with hands and feet and anteroposterior and lateral radiographs of the skull and spine, photographs and, whenever appropriate, autopsy, which is the most useful source of diagnostic information [42]. The pathologist should collect cartilage and bone, ideally femora and humeri, for histomorphological analysis, and fibroblasts, cartilage, bone or DNA should be stored for molecular study, given the recurrence risk [42]. In ECO, autopsy made it possible to document the holoprosencephaly, pituitary agenesis and adrenal hypoplasia, which ultrasound does not show [11]. Skeletal dysplasia registries (International Skeletal Dysplasia Registry, European Skeletal Dysplasia Network) and qualified centres should be used when the diagnosis is uncertain [42].

5.4. Genetic tests

The 2021 ACMG guideline, based on the GRADE method (Grading of Recommendations Assessment, Development and Evaluation), strongly recommends considering exome or genome sequencing as a first- or second-tier test in paediatric patients with congenital anomalies with onset before 1 year of age, because of its higher yield than conventional tests [16]. In the short-rib dysplasias, genetic diagnosis is usually made by targeted exome with ciliary gene panels, and when no variant, or only one, is found, as yet unknown genes should be considered [17]. In a cohort of 152 families from the International Skeletal Dysplasia Registry with SRPS, Jeune SRTD or Ellis-van Creveld syndrome, the exome identified causal variants in 120 (79%), in 14 genes, mainly DYNC2H1; CILK1 was represented by a single family in previous publications from the same registry [44]. The probability of a CILK1 variant in an unselected short-rib dysplasia is thus low, and the gene is only found if the panel includes it or if the analysis is exome or genome; the 2023 Nosology includes it in the skeletal ciliopathy group [3]. The composition of commercial panels was not assessed by the sources consulted. In the Old Order Amish community, targeted testing for the c.815G>A variant is a logical first step, by analogy with the targeted testing for the EVC founder variant NM_153717.3:c.1886+5G>T that GeneReviews recommends for Ellis-van Creveld syndrome in this population [10, 11].

5.5. Limitations of the tests

Exome sequencing detects missense, nonsense and canonical splice-site variants and small insertions or deletions, but does not reliably detect exon deletions or duplications, deep intronic or regulatory variants, or complex structures, and a negative result does not exclude the disease [10]. In the case of CILK1, ClinVar records a splice variant upstream of the first coding exon (NM_014920.5:c.-172-1G>A), located outside the regions usually covered by some panels [38]. Large 6p12 deletions that include the gene are detectable by microarray or copy number analysis, and in a recessive context a deletion in trans with a point variant would be compatible with the disease, a scenario that has not been published [38]. Interpretation of missense variants requires caution, because ECO is rare and ClinVar contains more than 300 records for the gene, many of uncertain significance [38].

5.6. Report content and interpretation

The report should state the gene (CILK1, with the previous symbol ICK), the transcript with version (NM_014920.5), the HGVS complementary DNA and protein nomenclature, zygosity, phase confirmed by parental testing and the ACMG/AMP class. The historical nomenclature “c.1305G>A” for p.(Arg272Gln) does not correspond to the current transcript and should be converted to c.815G>A [13, 27]. A single heterozygous CILK1 variant in a fetus with skeletal dysplasia does not explain the phenotype, and a rare heterozygous variant in a patient with JME should be reported in the context of the disputed evidence, without presenting it as causal: in ClinVar the EJM10 variants are listed as of uncertain significance or as risk factors [38, 39].

5.7. Practical algorithm

5.8. Prenatal and preimplantation diagnosis

When the familial variants have been identified, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic diagnosis are technically possible, as in any autosomal recessive disease with known variants [42]. The 2009 guideline recommends offering molecular testing in pregnancies at risk of homozygosity or compound heterozygosity and, when ultrasound suggests lethality, basing counselling on the clinical findings and considering molecular testing for confirmation [42]. In the absence of known variants, second-trimester ultrasound detected the anomalies in the Amish pregnancies and at 16 weeks in the Turkish fetus [11, 13].

6. Differential Diagnosis

6.1. Comparative table

6.2. Short rib-polydactyly syndromes and short-rib thoracic dysplasias

The SRPS and SRTD form a spectrum of autosomal recessive skeletal ciliopathies with a constricted thoracic cage, short ribs, short tubular bones and a trident acetabular roof, with variable polydactyly and visceral anomalies that differ between forms; some are lethal in the neonatal period from respiratory failure and others are compatible with life [46]. OMIM numbers them from SRTD1 to SRTD22: SRTD2 due to IFT80, SRTD3 due to DYNC2H1, SRTD4 due to TTC21B, SRTD5 due to WDR19, SRTD6 due to NEK1, SRTD7 due to WDR35, SRTD8 due to WDR60, SRTD9 due to IFT140, SRTD10 due to IFT172, SRTD11 due to WDR34, SRTD12 (Beemer-Langer), SRTD13 due to CEP120, SRTD14 due to KIAA0586, SRTD15 due to DYNC2LI1, SRTD16 due to IFT52, SRTD17 due to TCTEX1D2, SRTD18 due to IFT43, SRTD19 due to IFT81, SRTD20 due to INTU, SRTD21 due to KIAA0753 and SRTD22 due to FGF4; SRTD1 is only mapped to 15q13 [46]. ECO has not been given an SRTD number in OMIM, but the 2023 Nosology places it immediately after the SRPS, in the same group [1, 3]. Among the SRPS, type I (Saldino-Noonan) is the most severe, with irregular metaphyses with longitudinal spurs and vertebral ossification defects, and type II (Majewski) is characterised by short ovoid tibiae, median cleft lip and polysyndactyly [24]. The Turkish fetus with ECO had precisely a skeleton similar to that of SRTD type II, and the patient with p.(Glu80Lys) was included by the authors in an SRTD/SRPS cohort, which shows that, without molecular study, ECO is confused with these entities [13, 15]. Lahiry and colleagues stressed that the short ribs of Majewski and the heart defects of Mohr-Majewski were not present in the Amish patients, and Smith's mentions the occurrence of brain anomalies (pachygyria, small vermis, ventriculomegaly, absent olfactory bulbs) only occasionally in SRPS type II [11, 24]. The tables of Bone Dysplasias and of the Nosology show the heterogeneity: brain anomaly is present in Jeune SRTD, Saldino-Noonan, OFD IV and Beemer-Langer, and cleft lip and palate in Majewski, OFD IV and Beemer-Langer [3, 17]. In non-lethal SRTD, renal, hepatic and retinal complications arise in about 20% of survivors and increase with age [17].

6.3. Hydrolethalus syndrome

Hydrolethalus syndrome, described in 1981 mainly in Finnish families, shares with ECO prenatal-onset hydrocephalus, absence of the corpus callosum and septum pellucidum, occasional hypothalamic hamartoma or absent pituitary, micrognathia, cleft lip or palate, polydactyly and genital anomalies [47]. It is distinguished by the constant keyhole-shaped defect of the skull base, heart disease in half of cases (usually atrioventricular canal), abnormal lung lobation and the absence of severe micromelia and short ribs [47]. Birth prevalence in Finland is 5 per 100,000 [49]. The disease is caused by biallelic variants in HYLS1 (HLS1) and KIF7 (HLS2), also ciliary genes [48]. Lahiry and colleagues also discussed the “Majewski-hydrolethalus” picture, with short limbs and ribs and abnormal tibiae associated with hydramnios, hydrocephalus and a keyhole-shaped occipital bone [11].

6.4. Oral-facial-digital syndromes and Meckel-Gruber syndrome

The distinction between Majewski-type SRPS and oral-facial-digital syndrome IV (Mohr-Majewski) is difficult and requires molecular study; in OFD IV the ribs are thin and distorted but not short, survival beyond the neonatal period is possible and the tibia is less affected [17]. The orofacial findings of the Turkish fetus with ECO (multiple frenula, labiogingival fusion, tongue fused to the palate, natal teeth) bring it close to this group [13]. Meckel-Gruber syndrome combines occipital encephalocele, large multicystic kidneys, hepatic fibrosis and postaxial polydactyly [23].

6.5. Chromosomal causes and holoprosencephaly

Trisomy 13 and pseudotrisomy 13 (MIM 264480), with holoprosencephaly, hydrocephalus, polydactyly, heart disease and trisomy 13 facies without a chromosomal abnormality, enter the differential of ECO, and karyotype or microarray should precede or accompany the exome [11]. The absence of heart disease in the Amish patients with ECO is a feature that the authors used to distinguish them [11].

6.6. Juvenile myoclonic epilepsy

JME is genetically heterogeneous: OMIM lists susceptibility loci with identified genes (EJM5 due to GABRA1, EJM6 due to CACNB4, EJM8 due to CLCN2, EJM10 due to CILK1) and loci identified only by linkage (EJM2, EJM3, EJM4, EJM9), and has reclassified as a polymorphism the GABRD variant that defined EJM7 [50]. JME accounts for up to 26% of idiopathic generalised epilepsies [50].

7. Treatment and Risk Management Strategies

7.1. Principles

There is no treatment targeted at ECO and no clinical guideline specific to this syndrome or to the CILK1-related ciliopathies. Management rests on early prenatal diagnosis, counselling on prognosis, shared decision-making on continuation of the pregnancy, planning of delivery and perinatal and palliative care, according to the general recommendations for fetal skeletal dysplasias [42]. For CED6 the measures specific to cranioectodermal dysplasia apply, and for EJM10 the treatment of JME.

7.2. Ongoing pregnancy with a diagnosis of ECO

When ultrasound suggests a lethal dysplasia, the conclusions should be communicated to the attending physicians and to the pregnant woman, and counselling is based on the clinical findings, with molecular testing for confirmation [42]. If the pregnancy continues to a viable gestational age, the guideline recommends starting before delivery a joint consultation of obstetrics, neonatology, anaesthesia and genetics and defining a care plan [42]. In the ECO reports, severe hydrocephalus required drainage of cerebrospinal fluid from the ventricles to allow vaginal delivery, and a newborn without spontaneous breathing was intubated, extubated the following day and taken home for palliative care [11]. These approaches are case descriptions and not recommendations; the sources consulted do not compare delivery or ventilatory support strategies in ECO. When the pregnancy ends before viability, radiological confirmation and autopsy with tissue and DNA collection are recommended [42].

7.3. Liveborn newborn

Ventilatory support for respiratory failure due to pulmonary hypoplasia is the immediate measure, and its continuation depends on the diagnosis, the prognosis communicated and the parents' decision [15, 42]. The hypoplasia or agenesis of the pituitary and adrenals documented at autopsy justifies, in a newborn with ECO who survives, assessment of blood glucose, cortisol and pituitary function according to neonatology protocols; the sources describe no hormone measurements or replacement therapy in patients with ECO [11]. In the non-lethal forms of the short-rib spectrum, thoracic expansion surgery improves respiratory function and allows survival even with very restrictive thoracic cages, but no child with ECO has survived long enough for this option to be considered [17, 51].

7.4. Cranioectodermal dysplasia 6

Treatment of CED6 includes surveillance and treatment of renal and liver disease, infections and growth abnormalities; the specific treatment of CED is that of the craniosynostosis (craniotomy), infections and renal disease [14, 17].

7.5. Juvenile myoclonic epilepsy 10

There is no specific treatment for JME associated with CILK1 variants, and the sources consulted do not describe drug response in patients with these variants. Treatment is that of JME, which NICE (National Institute for Health and Care Excellence) guideline NG217, published on 27 April 2022 and last updated on 5 August 2026, includes among the idiopathic generalised epilepsies [18]. The table summarises the applicable recommendations.

The NICE committee considered that sodium valproate is the most effective drug in the idiopathic generalised epilepsies, that lamotrigine is associated with fewer adverse effects leading to discontinuation and with better quality of life, and that long-term valproate treatment reduces bone mineral density and increases the risk of osteomalacia, so supplementation should be considered in people at risk [18]. The guideline incorporates the safety measures of the MHRA (Medicines and Healthcare products Regulatory Agency): do not start valproate for the first time in people of either sex under 55 years of age, unless two specialists independently document that there is no other effective and tolerated treatment or that the reproductive risks do not apply; in women and girls of childbearing potential, use it only if other options have failed, after discussing the risks to the fetus and with the pregnancy prevention programme; in boys and men, recommend effective contraception (condom and contraception for the partner) during treatment and for 3 months afterwards, and discuss fertility risks with those planning to have children in the following year [18]. The doses, formulations and adjustments of these drugs are not specific to EJM10 and follow the general recommendations for the genetic epilepsies and the summaries of product characteristics (SmPC).

7.6. Regulatory status

The ClinicalTrials.gov search on 31 October 2026 returned no studies for “CILK1”, “endocrine-cerebro-osteodysplasia” or “short rib polydactyly” [52]. The absence of records in this database does not show that no other processes exist, and the regulatory orphan designation databases were not consulted for this entity.

7.7. Clinical trials and studies

There are no interventional trials in ECO, CED6 or EJM10. The study listed is genetic and observational, does not test treatments and is not specific to CILK1.

7.8. Preclinical research

In mice with a Cilk1 mutation, pharmacological activation of the SHH pathway by a Smoothened agonist (SAG) administered during gestation corrected several congenital defects, including cleft palate; the authors propose prenatal modulation of the hedgehog pathway as a possible strategy in ciliopathies, but the study is exclusively in animals [34]. In a Mak model of retinitis pigmentosa, gene delivery of Ick and inhibition of FGF receptors (negative regulators of Ick) attenuated retinal degeneration, pointing to CILK1 activation as a target for MAK retinitis pigmentosa, not for ECO [31]. In polycystic kidney disease due to deletion of Cilk1 in the collecting ducts, reduction of the KLC3 protein halted cyst progression in mice [35]. None of these approaches has been tested in humans.

7.9. Developments in the last 6–12 months

On 26 November 2025, the ClinGen expert panel for cystic kidney diseases and ciliopathies classified as “definitive” the validity of the association between CILK1 and autosomal recessive ciliopathy, grouping ECO and CED6 and leaving EJM10 for separate assessment (level: evidence curation) [4]. The description of CED6 in 2025 extended the spectrum to a form with survival (level: case series) [25]. In 2025, the demonstration that KATNIP variants associated with Joubert syndrome do not activate CILK1 linked the kinase to another ciliopathy (level: preclinical) [30], and in 2025 and 2026 mouse Cilk1 deletion models showed roles in dental and intestinal development (level: preclinical) [36, 37]. ClinVar received submissions of likely pathogenic variants for ECO without an associated publication (level: laboratory submission) [38]. There were no clinical trials, regulatory applications or authorisations.

7.10. Special populations

In a pregnancy with a fetus affected by ECO, severe hydrocephalus and polyhydramnios may determine the mode of delivery, and the legal framework for termination of pregnancy depends on the jurisdiction (section 8). In women with JME and childbearing potential, the choice of antiepileptic drug is subject to the restrictions on valproate and topiramate described in section 7.5 [18]. The sources do not describe pregnancy in women carrying CILK1 variants with JME.

7.11. Guideline recommendations and divergences

The applicable recommendations come from general documents: the 2009 guideline for the prenatal diagnosis of skeletal dysplasias [42], the 2021 ACMG guideline on exome and genome sequencing in congenital anomalies [16] and NICE NG217 for JME [18]. There is no divergence between societies on ECO, for lack of specific documents. In EJM10, the divergence concerns evidence and not recommendations: OMIM keeps the susceptibility entry, the original article attributes 7% of the cohort cases to CILK1 variants, an independent mouse study did not reproduce the epileptic phenotype and ClinVar classifies the variants as of uncertain significance, while ClinGen has not yet published its assessment of this association [4, 7, 12, 38, 39].

8. Reproductive Options and Family Counselling

8.1. Recurrence risk in ECO

ECO is autosomal recessive: when both parents are heterozygous, each pregnancy has a 25% chance of an affected fetus, 50% of a carrier and 25% of a non-carrier [10, 11]. In the published families, the parents were heterozygous and unaffected, and in the Amish family there were several affected fetuses in two sibships linked by consanguinity [11, 13]. Heterozygous carriers of ECO variants have, according to the sources, no described manifestations; the possibility of susceptibility to epilepsy in carriers, suggested by the association with EJM10, has not been studied in families with ECO and is not established [12, 39].

8.2. Carrier testing and cascade testing

Once the variants of the index case have been identified, parental testing confirms phase, and testing of adult siblings and partners allows the risk for future pregnancies to be estimated. In the Old Order Amish community, the carrier frequency of c.815G>A was 5 in 257 healthy controls (about 1 in 51), and the probability that two unrelated members of the community are both carriers is therefore appreciable, unlike outside it, where the variant was not found in 2855 individuals [11]. In consanguineous couples of other origins, as in the Turkish family, the risk depends on the familial variant [13]. In minors, carrier testing without immediate medical benefit should be deferred until the age of autonomous decision-making, according to the general principles of clinical genetics.

8.3. Prenatal and preimplantation diagnosis

With the familial variants known, prenatal diagnosis by first-trimester chorionic villus sampling or by amniocentesis, and preimplantation genetic diagnosis, make it possible to identify affected fetuses or embryos; the 2009 guideline recommends offering molecular testing in at-risk pregnancies and identifying the parents' variants, ideally before pregnancy [42]. Second-trimester ultrasound detected the anomalies in all the Amish pregnancies in which it was performed, and the 16-week scan showed them in the Turkish fetus [11, 13].

8.4. Legal framework for termination of pregnancy in Portugal

In Portugal, article 142 of the Penal Code (as amended by Law no. 16/2007) declares non-punishable a termination of pregnancy performed by a physician, in an official or officially recognised health establishment and with the consent of the pregnant woman, when there are sound reasons to foresee that the unborn child will suffer, incurably, from a serious disease or congenital malformation, if performed within the first 24 weeks, except for non-viable fetuses, in which it may be performed at any time [53]. The circumstances are certified in a written and signed statement before the procedure by a physician other than the one performing it, and consent is given in a signed document, whenever possible at least three days in advance [53]. The sources consulted do not include Brazilian legislation.

8.5. Juvenile myoclonic epilepsy 10

EJM10 is catalogued as autosomal dominant with incomplete penetrance: the risk of transmitting a heterozygous variant is 50% in each pregnancy, but the probability of disease in a carrier is unknown, because in the families described there were carriers without epilepsy and relatives with EEG changes only [7, 12]. Given the classification of the variants as of uncertain significance and the failure to reproduce the phenotype in mice, predictive testing of asymptomatic relatives has no demonstrated clinical utility [38, 39]. As for the empirical risk of epilepsy in first-degree relatives of a patient with idiopathic generalised epilepsy, without considering genotype, the Oxford Desk Reference cites a cumulative risk of 4.7% up to 40 years of age in first-degree relatives in a study of epilepsies, with higher concordance in monozygotic twins (0.62) than in dizygotic twins (0.18) [54]. A couple in which both partners carry CILK1 variants would, in addition, have a 25% risk of a child with a recessive form, but the severity would depend on the variants, which in EJM10 are different from those of ECO.

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency

Orphanet records 7 cases or families with ECO, based on the 2009 and 2016 articles, and classifies the prevalence as below 1 per 1,000,000 [2]. The publications describe three families and about nine affected individuals: six to seven Old Order Amish fetuses and newborns, a Turkish fetus and a newborn of unstated origin [11, 13, 15]. These numbers count published cases and do not measure the frequency of the disease; recent submissions to ClinVar suggest other unpublished cases [38]. For the neighbouring entities, Orphanet records 34 cases or families with Majewski-type SRPS, a birth prevalence of Jeune syndrome of 1.4 per 100,000 in Europe (EUROCAT, 2005–2012) and of 5 per 100,000 for hydrolethalus syndrome in Finland [49, 55, 56]. JME has a prevalence of 1.7 per 100,000 in a Norwegian study, affects about 8.5 million people worldwide according to the estimates cited by Bailey and colleagues, and accounts for up to 26% of idiopathic generalised epilepsies [9, 12, 50].

9.2. Founder populations

The c.815G>A p.(Arg272Gln) variant is private to the Old Order Amish community, with an allele frequency of 0.97% and an estimated frequency of homozygotes of about 1 in 10,000 in this population [11]. The Amish community has other founder variants for skeletal ciliopathies, such as the EVC variant NM_153717.3:c.1886+5G>T of Ellis-van Creveld syndrome in Lancaster County, Pennsylvania [10]. The Turkish family was consanguineous and the p.(Gly120Cys) variant was absent from 1092 Turkish exomes [13]. CED6 has a founder effect in the Iraqi Turkmen population of Tal Afar [25]. The EJM10 cohort was mainly of mixed European and Amerindian ancestry (17 of the 24 probands with variants), European (4) and Japanese (3), recruited at centres in the United States, Mexico, Honduras, Brazil and Japan; the main text does not state how many Brazilian patients had variants [12].

9.3. European, Portuguese and Brazilian data

The sources consulted include no European cases of ECO, nor data from Portugal or Brazil, which does not show that none exist. The GeneReviews chapter on Ellis-van Creveld syndrome, useful in the differential, is authored by a group from the Medical Genetics Department of the Centro Hospitalar Universitário de Santo António, in Porto [10]. The observational study NCT03400371 is recruiting patients with JME in the United Kingdom, Europe and North America [52].

9.4. European reference networks and patient associations

Cases of ECO and other skeletal ciliopathies fall within ERN ITHACA (rare malformation syndromes, intellectual disability and other neurodevelopmental disorders) and ERN BOND (rare bone diseases); the renal complications of the ciliopathies within ERKNet; and the rare epilepsies within EpiCARE [19–22]. Patient associations include the Ciliopathy Alliance, recommended by the Oxford Desk Reference for the ciliopathies, EURORDIS (European organisation of patients with rare diseases) and, for epilepsy in Portugal, the Liga Portuguesa Contra a Epilepsia [23, 57–59]. Skeletal dysplasia registries (International Skeletal Dysplasia Registry and European Skeletal Dysplasia Network) support the diagnosis of uncertain cases [42].

10. Illustrative Clinical Case

10.1. Turkish fetus with ECO and the p.(Gly120Cys) variant

In 2016 Oud and colleagues described the first child of a consanguineous Turkish couple [13]. The 16-week ultrasound showed cystic hygroma, scalp oedema, ascites, bilateral ventriculomegaly, very short tubular bones, polydactyly of the hands and feet and short ribs, and SRTD was suggested; at 19 weeks large hyperechogenic kidneys and a ventricular septal defect were added, and the fetal karyotype was 46,XY [13]. The pregnancy was terminated at 33 weeks. The fetus weighed 2450 g (90th–97th percentile), measured 41 cm (10th–25th percentile) and had a head circumference of 34 cm (above the 97th percentile) [13]. Post-mortem examination showed a hydropic appearance, a short thorax and very short limbs, with oral-facial-digital defects: high broad forehead, deep-set eyes, hypertelorism, a very small nose with anteverted nares, a median pseudocleft of the upper lip, soft palate cleft, natal teeth, labiogingival fusion, an irregular alveolar ridge, tongue fused to the lower palate, multiple frenula, micrognathia, low-set posteriorly rotated ears, postaxial heptadactyly of the hands and preaxial of the feet, sex reversal with hypoplastic labial folds and an anteriorly placed ectopic anus [13]. Radiographs showed very short tubular bones, torpedo-shaped femora, hypoplastic ovoid tibiae, premature epiphyseal ossification in the lower limbs, short ribs, slightly deficient vertebral ossification and small ilia of normal shape, similar to those of SRTD type II [13]. Homozygosity mapping delimited four candidate regions, including one of 21 Mb on chromosome 6, and exome sequencing of the fetus identified, among only five homozygous variants in the identical-by-descent blocks, the variant NM_014920.3:c.358G>T p.(Gly120Cys) in ICK, at a glycine residue conserved in all vertebrates, in the catalytic domain [13]. The variant was homozygous in the fetus, heterozygous in both parents and heterozygous or absent in 20 other relatives, and was not present in ExAC or in 1092 Turkish exomes [13]. On comparing the phenotype with that of the Amish fetuses, the authors concluded that this was the second family with ECO worldwide, although it was not possible to determine whether the fetus had the typical endocrine anomalies [13]. In transfected cells, the mutant protein accumulated at a bulging ciliary tip, and fibroblasts from an Amish patient showed reduced ciliogenesis [13]. Figure 1 of the article brings together the pedigree, the photographs and radiographs of the fetus (hydrops, tetramicromelia, narrow thorax, polydactyly with hypoplastic digits, cleft lip and palate, natal teeth, multiple frenula) and the variant map: https://pmc.ncbi.nlm.nih.gov/articles/PMC4827216/figure/Fig1/ [13]. Figure 2, with the ciliary localisation of the mutant proteins and ciliogenesis in fibroblasts, is at https://pmc.ncbi.nlm.nih.gov/articles/PMC4827216/figure/Fig2/ [13].

10.2. Lessons from the case

The case shows that ECO can present in the second trimester as a short-rib dysplasia with polydactyly, indistinguishable on ultrasound from Majewski-type SRPS, and that the presence of midline orofacial anomalies and ventriculomegaly in a fetus with this pattern justifies including CILK1 in the molecular analysis [13]. It also shows the variability of the syndrome: this fetus did not have holoprosencephaly, present in the four Amish assessed, and had a ventricular septal defect, absent in the Amish [13]. The termination of pregnancy and post-mortem study reported are the management of this case, in line with what the fetal skeletal dysplasia guideline recommends regarding diagnostic confirmation after delivery [42].

10.3. Newborn with an SRPS phenotype and the p.(Glu80Lys) variant

By contrast, the patient described by Paige Taylor and colleagues was identified on prenatal ultrasound with severe skeletal dysplasia, hydrocephalus, genital anomalies and craniofacial dysmorphism, was born at 38 weeks weighing 4050 g with Apgar scores of 2, 6 and 7, was intubated for respiratory failure and died within a few hours; he had a long narrow thorax with short ribs, postaxial polydactyly of all four limbs, fused eyelids and absent external genitalia, and was homozygous for NM_014920.5:c.238G>A p.(Glu80Lys), which abolishes kinase activity [15]. The radiographs are in Figure 1 of the article (https://pmc.ncbi.nlm.nih.gov/articles/PMC5291234/figure/ddw240-F1/), whose legend was consulted in the full text [15].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The prognosis of ECO is lethal: among the published affected individuals there was death on the third day of life, death during delivery, death hours after birth or termination of pregnancy for medical reasons between 23 and 33 weeks [11, 13, 15]. No survivors have been described, and the data do not allow survival with prolonged intensive care to be estimated. The p.(Arg272Gln) mouse model dies at birth from respiratory failure, with structural pulmonary hypoplasia, which supports the view that the cause of death does not depend solely on thoracic restriction [32]. CED6 has a different prognosis, with survival in childhood and renal and hepatic morbidity [14], and EJM10 has the prognosis of JME, in which about 16–17% of patients are drug-resistant according to the estimates cited by Bailey and colleagues [12].

11.2. Follow-up

In ECO there is no long-term follow-up of the patient; follow-up concerns the family, with genetic counselling after molecular confirmation, carrier testing in interested relatives and planning of subsequent pregnancies [42]. The table summarises the measures, including, for the rare non-lethal cases of the short-rib spectrum, the follow-up protocol proposed for Jeune syndrome.

The Jeune syndrome protocol, proposed by de Vries and colleagues and cited by Smith's Recognizable Patterns of Human Malformation, focuses the first two years on respiratory disease and has not been validated in ECO, in which there are no survivors [51]. Follow-up of CED6 follows the recommendations specific to cranioectodermal dysplasia.

12. Summary and Final Considerations

Endocrine-cerebro-osteodysplasia syndrome is an autosomal recessive skeletal ciliopathy caused by biallelic variants in the kinase domain of CILK1, the gene formerly named ICK, and was described in 2009 in an Old Order Amish family, in which the c.815G>A p.(Arg272Gln) variant has an allele frequency of 0.97% [11]. It combines midline brain malformations (holoprosencephaly, hydrocephalus, corpus callosum agenesis), median cleft lip and palate, micromelia with bowing, polydactyly, pituitary and adrenal hypoplasia and abnormal genitalia, and all published affected individuals died in the perinatal period or underwent termination of pregnancy for medical reasons [11, 13, 15]. The second family, from Turkey, and a newborn with an SRPS phenotype showed that ECO is confused with the short rib-polydactyly dysplasias, and the degree of loss of kinase activity has been proposed as the explanation for the difference between the two pictures [13, 15]. CILK1 regulates ciliary length and intraflagellar transport, and its loss alters GLI3 processing and hedgehog signalling, with consequences for the skeleton, lung and brain [15, 32, 33]. The same gene causes, through a distal truncating variant, cranioectodermal dysplasia 6, with survival, and has been associated in the heterozygous state with susceptibility to juvenile myoclonic epilepsy, an association that an independent mouse study did not reproduce and that ClinGen separated from the recessive form, classified as definitive [4, 12, 25, 39]. Diagnosis is made by exome, genome or a skeletal ciliopathy panel that includes the gene, with the current nomenclature NM_014920.5 instead of the historical notation, and management is that of lethal fetal skeletal dysplasias: assessment at a specialised centre, counselling, perinatal and palliative care, autopsy and reproductive counselling with prenatal or preimplantation diagnosis in subsequent pregnancies [16, 27, 42]. There are no treatments or clinical trials, and research is limited to animal models [34].

13. References

1. OMIM® Online Mendelian Inheritance in Man. #612651 ENDOCRINE-CEREBROOSTEODYSPLASIA; ECO. Baltimore (MD): Johns Hopkins University; last edited 09/09/2025. https://omim.org/entry/612651
2. Orphanet. Endocrine-cerebro-osteodysplasia syndrome. ORPHA:199332. Paris: INSERM; data updated 2026-06/07. https://www.orpha.net/en/disease/detail/199332
3. Unger S, Ferreira CR, Mortier GR, Ali H, Bertola DR, Calder A, et al. Nosology of genetic skeletal disorders: 2023 revision. Am J Med Genet A. 2023;191(5):1164-1209. doi:10.1002/ajmg.a.63132. PMID: 36779427. PMCID: PMC10081954.
4. ClinGen. CILK1 (HGNC:21219) — gene-disease validity: CILK1 ciliopathy, autosomal recessive, Definitive (Kidney Cystic and Ciliopathy Disorders GCEP, 26/11/2025). https://search.clinicalgenome.org/kb/genes/HGNC:21219
5. OMIM® Online Mendelian Inheritance in Man. *612325 CILIOGENESIS-ASSOCIATED KINASE 1; CILK1. Johns Hopkins University; last edited 08/10/2025. https://omim.org/entry/612325
6. GeneGuidance. ICK (CILK1) gene record. Version of 2026-09-26.
7. OMIM® Online Mendelian Inheritance in Man. #617924 EPILEPSY, JUVENILE MYOCLONIC, SUSCEPTIBILITY TO, 10; EJM10. Johns Hopkins University; last edited 07/07/2025. https://omim.org/entry/617924
8. Orphanet. Short rib-polydactyly syndrome. ORPHA:1505. https://www.orpha.net/en/disease/detail/1505
9. Orphanet. Juvenile myoclonic epilepsy. ORPHA:307. https://www.orpha.net/en/disease/detail/307
10. Da Silva JD, Tkachenko N, Soares AR. Ellis-van Creveld Syndrome. 2023 Oct 26. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK596643/
11. Lahiry P, Wang J, Robinson JF, Turowec JP, Litchfield DW, Lanktree MB, et al. A multiplex human syndrome implicates a key role for intestinal cell kinase in development of central nervous, skeletal, and endocrine systems. Am J Hum Genet. 2009;84(2):134-147. doi:10.1016/j.ajhg.2008.12.017. PMID: 19185282. PMCID: PMC2668000.
12. Bailey JN, de Nijs L, Bai D, Suzuki T, Miyamoto H, Tanaka M, et al. Variant intestinal-cell kinase in juvenile myoclonic epilepsy. N Engl J Med. 2018;378(11):1018-1028. doi:10.1056/NEJMoa1700175. PMID: 29539279.
13. Oud MM, Bonnard C, Mans DA, Altunoglu U, Tohari S, Ng AYJ, et al. A novel ICK mutation causes ciliary disruption and lethal endocrine-cerebro-osteodysplasia syndrome. Cilia. 2016;5:8. doi:10.1186/s13630-016-0029-1. PMID: 27069622. PMCID: PMC4827216.
14. OMIM® Online Mendelian Inheritance in Man. #621337 CRANIOECTODERMAL DYSPLASIA 6; CED6. Johns Hopkins University; last edited 08/10/2025. https://omim.org/entry/621337
15. Paige Taylor S, Kunova Bosakova M, Varecha M, Balek L, Barta T, Trantirek L, et al. An inactivating mutation in intestinal cell kinase, ICK, impairs hedgehog signalling and causes short rib-polydactyly syndrome. Hum Mol Genet. 2016;25(18):3998-4011. doi:10.1093/hmg/ddw240. PMID: 27466187. PMCID: PMC5291234.
16. Manickam K, McClain MR, Demmer LA, Biswas S, Kearney HM, Malinowski J, et al. Exome and genome sequencing for pediatric patients with congenital anomalies or intellectual disability: an evidence-based clinical guideline of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(11):2029-2037. doi:10.1038/s41436-021-01242-6. PMID: 34211152.
17. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Short rib (±polydactyly) dysplasias. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 380–408.
18. National Institute for Health and Care Excellence. Epilepsies in children, young people and adults. NICE guideline NG217. London: NICE; 2022 (last updated 5 August 2026). https://www.nice.org.uk/guidance/ng217
19. ERN ITHACA — European Reference Network on Rare Malformation Syndromes, Intellectual and Other Neurodevelopmental Disorders. https://ern-ithaca.eu
20. ERN BOND — European Reference Network on Rare Bone Diseases. https://ernbond.eu
21. EpiCARE — European Reference Network for Rare and Complex Epilepsies. https://epi-care.eu
22. ERKNet — European Rare Kidney Disease Reference Network. https://www.erknet.org
23. Firth HV, Hurst JA. Ciliopathies. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 366–369.
24. Jones KL, Jones MC, del Campo M. Short rib–polydactyly syndromes. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 474–477.
25. Sezer A, Oner SS, Saat H, Turan MG, Gungor T, Cevik S, et al. A homozygous frameshift variant in the CILK1 gene causes cranioectodermal dysplasia. Eur J Hum Genet. 2025;33(10):1240-1251. doi:10.1038/s41431-025-01902-0. PMID: 40615527. PMCID: PMC12479750.
26. Fu Z, Gailey CD, Wang EJ, Brautigan DL. Ciliogenesis associated kinase 1: targets and functions in various organ systems. FEBS Lett. 2019;593(21):2990-3002. doi:10.1002/1873-3468.13600. PMID: 31506943. PMCID: PMC6848764.
27. 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):61-68. doi:10.1002/humu.23348. PMID: 28967166.
28. Wang EJ, Gailey CD, Brautigan DL, Fu Z. Functional alterations in ciliogenesis-associated kinase 1 (CILK1) that result from mutations linked to juvenile myoclonic epilepsy. Cells. 2020;9(3):694. doi:10.3390/cells9030694. PMID: 32178256. PMCID: PMC7140639.
29. Turner JS, McCabe EA, Kuang KW, Gailey CD, Brautigan DL, Limerick A, et al. The scaffold protein KATNIP enhances CILK1 control of primary cilia. Mol Cell Biol. 2023;43(9):472-480. doi:10.1080/10985549.2023.2246870. PMID: 37665596. PMCID: PMC10512882.
30. Carpenter EH, Chu CY, Limerick A, Brautigan DL, Fu Z. Human disease variants of KATNIP fail to support CILK1 activation and control of primary cilia. J Cell Sci. 2025;138(20):jcs264056. doi:10.1242/jcs.264056. PMID: 40621737. PMCID: PMC12624974.
31. Chaya T, Maeda Y, Tsutsumi R, Ando M, Ma Y, Kajimura N, et al. Ccrk-Mak/Ick signaling is a ciliary transport regulator essential for retinal photoreceptor survival. Life Sci Alliance. 2024;7(11):e202402880. doi:10.26508/lsa.202402880. PMID: 39293864. PMCID: PMC11412320.
32. Tong Y, Park SH, Wu D, Xu W, Guillot SJ, Jin L, et al. An essential role of intestinal cell kinase in lung development is linked to the perinatal lethality of human ECO syndrome. FEBS Lett. 2017;591(9):1247-1257. doi:10.1002/1873-3468.12644. PMID: 28380258. PMCID: PMC5466854.
33. Ding M, Jin L, Xie L, Park SH, Tong Y, Wu D, et al. A murine model for human ECO syndrome reveals a critical role of intestinal cell kinase in skeletal development. Calcif Tissue Int. 2018;102(3):348-357. doi:10.1007/s00223-017-0355-3. PMID: 29098359. PMCID: PMC5820141.
34. Shin JO, Song J, Choi HS, Lee J, Lee K, Ko HW, et al. Activation of sonic hedgehog signaling by a Smoothened agonist restores congenital defects in mouse models of endocrine-cerebro-osteodysplasia syndrome. EBioMedicine. 2019;49:305-317. doi:10.1016/j.ebiom.2019.10.016. PMID: 31662288. PMCID: PMC6945271.
35. Rah G, Cha H, Kim J, Song J, Kim H, Oh YK, et al. KLC3 regulates ciliary trafficking and cyst progression in CILK1 deficiency-related polycystic kidney disease. J Am Soc Nephrol. 2022;33(9):1726-1741. doi:10.1681/ASN.2021111455. PMID: 35961787. PMCID: PMC9529174.
36. Song J, Je S, Lee B, Ko HW. Cilk1 is essential for mesenchymal cilia maintenance and epithelial-mesenchymal crosstalk in intestinal villus morphogenesis. Cell Mol Gastroenterol Hepatol. 2026;20(3):101672. doi:10.1016/j.jcmgh.2025.101672. PMID: 41192595. PMCID: PMC12804004.
37. Kyeong M, Jeong JK, Adasooriya D, Kan S, Kim J, Song J, et al. Progressive tooth pattern changes in Cilk1-deficient mice depending on Hedgehog signaling. Int J Oral Sci. 2025;17(1):71. doi:10.1038/s41368-025-00405-4. PMID: 41320695. PMCID: PMC12665794.
38. National Center for Biotechnology Information. ClinVar: CILK1[gene] (328 records). https://www.ncbi.nlm.nih.gov/clinvar/?term=CILK1%5Bgene%5D
39. Salvati KA, Mason AJ, Gailey CD, Wang EJ, Fu Z, Beenhakker MP. Mice harboring a non-functional CILK1/ICK allele fail to model the epileptic phenotype in patients carrying variant CILK1/ICK. Int J Mol Sci. 2021;22(16):8875. doi:10.3390/ijms22168875. PMID: 34445580. PMCID: PMC8396347.
40. Limerick A, McCabe EA, Turner JS, Kuang KW, Brautigan DL, Hao Y, et al. An epilepsy-associated CILK1 variant compromises KATNIP regulation and impairs primary cilia and Hedgehog signaling. Cells. 2024;13(15):1258. doi:10.3390/cells13151258. PMID: 39120290. PMCID: PMC11311665.
41. Hirsch E, French J, Scheffer IE, Bogacz A, Alsaadi T, Sperling MR, et al. ILAE definition of the idiopathic generalized epilepsy syndromes: position statement by the ILAE Task Force on Nosology and Definitions. Epilepsia. 2022;63(6):1475-1499. doi:10.1111/epi.17236. PMID: 35503716.
42. Krakow D, Lachman RS, Rimoin DL. Guidelines for the prenatal diagnosis of fetal skeletal dysplasias. Genet Med. 2009;11(2):127-133. doi:10.1097/GIM.0b013e3181971ccb. PMID: 19265753. PMCID: PMC2832320.
43. Pajkrt E, Chitty LS. A sonographic approach to the prenatal diagnosis of skeletal dysplasias. Prenat Diagn. 2019;39(9):701-719. doi:10.1002/pd.5501. PMID: 31173381.
44. Zhang W, Taylor SP, Ennis HA, Forlenza KN, Duran I, Li B, et al. Expanding the genetic architecture and phenotypic spectrum in the skeletal ciliopathies. Hum Mutat. 2018;39(1):152-166. doi:10.1002/humu.23362. PMID: 29068549. PMCID: PMC6198324.
45. OMIM® Online Mendelian Inheritance in Man. %269860 SHORT-RIB THORACIC DYSPLASIA 12; SRTD12. Johns Hopkins University; last edited 03/15/2022. https://omim.org/entry/269860. Accessed 31 October 2026.
46. OMIM® Online Mendelian Inheritance in Man. %208500 SHORT-RIB THORACIC DYSPLASIA 1 WITH OR WITHOUT POLYDACTYLY; SRTD1. Baltimore (MD): Johns Hopkins University; last edited 07/02/2025. https://omim.org/entry/208500. Accessed 31 October 2026.
47. Jones KL, Jones MC, del Campo M. Hydrolethalus syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 254–255.
48. OMIM® Online Mendelian Inheritance in Man. #236680 HYDROLETHALUS SYNDROME 1; HLS1. Johns Hopkins University; last edited 01/17/2024. https://omim.org/entry/236680. Accessed 31 October 2026.
49. Orphanet. Hydrolethalus. ORPHA:2189. https://www.orpha.net/en/disease/detail/2189
50. OMIM® Online Mendelian Inheritance in Man. #254770 EPILEPSY, MYOCLONIC JUVENILE; EJM. Johns Hopkins University; last edited 06/28/2024. https://omim.org/entry/254770. Accessed 31 October 2026.
51. Jones KL, Jones MC, del Campo M. Jeune thoracic dystrophy (asphyxiating thoracic dystrophy). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 480–481.
52. King's College London. Biology of Juvenile Myoclonic Epilepsy (BIOJUME). ClinicalTrials.gov NCT03400371. https://clinicaltrials.gov/study/NCT03400371
53. Portugal. Código Penal, artigo 142.º (Interrupção da gravidez não punível), as amended by Lei n.º 16/2007, de 17 de abril. https://www.pgdlisboa.pt/leis/lei_mostra_articulado.php?artigo_id=109A0142&nid=109&tabela=leis
54. Firth HV, Hurst JA. Epilepsy in adults. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 416–419.
55. Orphanet. Short rib-polydactyly syndrome, Majewski type. ORPHA:93269. https://www.orpha.net/en/disease/detail/93269
56. Orphanet. Jeune syndrome. ORPHA:474. https://www.orpha.net/en/disease/detail/474
57. Ciliopathy Alliance [Internet]. London [accessed 2026-10-31]. Available from: https://www.ciliopathyalliance.org
58. EURORDIS — Rare Diseases Europe [Internet]. Paris [accessed 2026-10-31]. Available from: https://www.eurordis.org
59. Liga Portuguesa Contra a Epilepsia [Internet]. Porto [accessed 2026-10-31]. Available from: https://www.epilepsia.pt