Craniofacial and anthropometric phenotype in ankyloblepharon–ectodermal defects–cleft lip/palate syndrome (Hay–Wells syndrome) in a cohort of 17 patients


  • How to cite this article: Sutton VR, Plunkett K, Dang DX, Lewis RA, Bree AF, Bacino CA. 2009. Craniofacial and anthropometric phenotype in ankyloblepharon–ectodermal defects–cleft lip/palate syndrome (Hay–Wells syndrome) in a cohort of 17 patients. Am J Med Genet Part A 149A:1916–1921.


Ankyloblepharon–ectodermal dysplasia–cleft lip/palate (AEC) syndrome and Rapp–Hodgkin syndrome are well-characterized clinical entities caused by mutations in the TP63 gene. While AEC and Rapp–Hodgkin had been thought to be clinically distinct entities, the elucidation of their molecular etiology confirmed that they are a clinical continuum as opposed to distinct disorders. We have evaluated 17 patients with AEC syndrome using a systematic clinical approach. In our study, we have identified new features and others that were thought to occur only rarely. These include short stature and poor weight gain with preservation of head circumference in nearly all subjects, trismus in 35% and hypospadias in 78% of males. In addition, we describe the frequency of phenotypic features and demonstrate the extreme clinical variability in the largest cohort of AEC individuals reported in the literature thus far. © 2009 Wiley-Liss, Inc.


Ankyloblepharon–ectodermal defect–cleft lip/palate (AEC) syndrome, also known as Hay–Wells syndrome, was initially described in seven individuals from four unrelated families who presented with ankyloblepharon (fusion of the lateral margins of the lids), atresia of the lacrimal ducts or absence of lacrimal puncta, absent eyelashes, coarse wiry hair, pili torti, nail dystrophy, cleft palate, cleft lip, dental abnormalities, including hypodontia or anodontia, maxillary hypoplasia, and mild hypohydrosis [Hay and Wells, 1976]. The clinical features are variable with 50% reported to have clefting, and only 30% with ankyloblepharon filiforme adnatum [Rapp and Hodgkin, 1968; Hay and Wells, 1976; Greene et al., 1987]. The features of ectodermal dysplasia are more prevalent than the previous two anomalies; although, the true incidence unclear.

Prior to the description of AEC by Hay and Wells, Rapp and Hodgkin reported an autosomal dominant form of dyshidrotic ectodermal dysplasia, in association with cleft lip and palate, in one single family. Three individuals from this described family—mother, daughter, and son—were equally affected. Skin biopsies of the affected individuals showed paucity/absence of sweat glands [Rapp and Hodgkin, 1968]. Further reports of similar associations of ectodermal dysplasia and clefting quickly followed and the syndrome has been since known as Rapp–Hodgkin ectodermal dysplasia or Rapp–Hodgkin syndrome (RHS) [Crawford et al., 1989; Felding and Bjorklund, 1990; Breslau-Siderius et al., 1991; O'Donnell and James, 1992]. Later reports speculate about the possibility that AEC and RHS are not clinically distinct entities [Cambiaghi et al., 1994]. This was confirmed when molecular analysis revealed that both entities were caused by mutations in the sterile α-motif (SAM) and transactivation inhibitory (TI) domains of the TP63 gene (previously known as TP73L, TP53L, and TP53CP) on 3q27 [McGrath et al., 2001; Tsutsui et al., 2003; Shotelersuk et al., 2005]. TP63 shares significant homology with p53, a well-known tumor suppressor gene.

The TP63 (tumor protein TP63) gene has four domains including a TI domain, a DNA binding domain, an oligomerization domain, in addition to a carboxy-terminal SAM domain. The mutations for AEC and RHSs are localized to the SAM and TI domains of the molecule [Kantaputra et al., 2003; Chan et al., 2005]. SAM domains are found in a large number of proteins and are important players in cell signaling from tyrosine kinase receptors, transcription factors, serine/threonine kinases, adapter proteins, polyhomeotic proteins, and liprins [Thanos and Bowie, 1999; Thanos et al., 1999].

The TP63 gene is a known regulator of epithelial development/differentiation. Mutations in this gene have also been found in other disorders of limb development. Mutations in the DNA-binding domain of TP63 cause ectodermal dysplasia and facial clefting (EEC) and split hand/foot malformation (SHFM). Limb–mammary syndrome (LMS) and acro-dermato-ungual-lacrimal-tooth (ADULT) syndrome are both rare disorders that appear to be caused by specific mutations in TP63 different from those seen in other disorders [Barrow et al., 2002; Brunner et al., 2002a,b; Rinne et al., 2006].

The mechanisms by which TP63 mutations cause disease are still unclear; although, it is likely the result of a dominant-negative and/or gain of function mutation rather than loss of function [Brunner et al., 2002b]. It is believed that TP63 keeps a dual role essential for the development and the maintenance of the epidermis allowing the initiation of epithelial stratification. The TP63 gene is mainly expressed in the basal compartment of the epidermis and normally requires down-regulation for terminal differentiation [Koster et al., 2004, 2007]. Recent evidence suggests that TP63 mutations in the SAM domain interact with the regulation of the fibroblast growth factor receptor gene FGFR-2 and mediate alternate mRNA splicing and Notch signaling. This process may actually affect multiple other signaling pathways including the expression levels of Jagged 1, Jagged 2, Notched 1, Notched 2, Hey 1, and Hey 2 [Fomenkov et al., 2003; Huang et al., 2005].

We have recently had the opportunity to perform a detailed examination of 17 individuals from 13 unrelated families with AEC syndrome as part of the AEC International Research Symposium. No larger cohorts of individuals with AEC syndrome have been studied and reported to date.


We have recently studied and examined a total of 17 individuals from 13 unrelated families who were participating at the International Research Symposium for AEC syndrome, sponsored by the National Foundation for Ectodermal Dysplasias (NFED). All of these individuals and/or their families consented to participate in an IRB-approved research protocol allowing clinical evaluations and photographs. A full clinical evaluation was performed including a number of anthropometric measurements. The age of the participants ranged from 4 months to 30 years. From the 17 subjects studied, 9 were males and 8 were females. Patient measurements were performed manually by one of two examiners. z-Scores were calculated for each measurement according to age and sex using the ABase program, version 1.2 ( [Zankl et al., 2002; Zankl and Molinari, 2003]. Statistical calculations were made on STATA software (1984–2003). Pictures were obtained of each individual.

Clinical Evaluations

All patients were simultaneously examined by one of two experienced clinical geneticists. A complete family history was also obtained as part of this evaluation. A full clinical exam in each individual was targeted to evaluate previously known features in AEC syndrome, as well as general clinical parameters. Among the clinical features evaluated, we recorded the following anthropometrics: height, weight, fronto-occipital circumference (FOC). We looked for the presence or absence of: scalp erosions, plamar and plantar keratoderma, hyperkeratosis, hyperpigmentation, ankyloblepharon filiforme adnatum, lacrimal duct atresia, sparse eyelashes, sparse lateral eyebrows, conjunctivitis, blepharitis, ear length, maxillary hypoplasia, oval face, narrow nose, broad nasal bridge, hypoplasia of alae nasi, philtrum length, thin vermillion border, microstomia, trismus, small mandible, atretic external auditory canal, cupped auricles, cleft lip and or cleft palate, partial anodontia or hypodontia, conical teeth, wide spaced teeth, hypospadias in males, syndactyly of hands and feet, ectrodactyly, keratitis, nail changes, pterygia, alopecia, and general intelligence. We also recorded the presence of hair abnormalities: wiry and sparse hair, patchy alopecia, sparse body hair. Limb abnormalities recorded included ectrodactyly and 2–3 toe soft tissue syndactyly. The chest was evaluated for the presence of supernumerary nipples. Males were examined for micropenis and hypospadias.

Regarding the eye exam, a single observer (RAL) performed a thorough examination of the external, adnexal, and anterior ocular structures with subject- and age-specific standard equipment, including rulers, tapes, or calipers, and a portable binocular slitlamp biomicroscope (Zeiss Model SLO-10). Measurements of the intermedial canthal, interpupillary, and interlateral canthal distances and the horizontal corneal diameters were recorded in the usual clinical fashion. Any variations in the usual anatomic structures of the brows, lids and skin, lashes, puncta, and caruncles, inspected under magnification, were also documented. As well, historical data regarding presence of ankyloblepharon was also obtained since this is typically a transient finding in the newborn period. The age-corrected midfacial dimensions were compared to standard normative databases [Zankl et al., 2002; Zankl and Molinari, 2003].

Molecular Studies

As previously discussed, mutations in the 5′ tail end of the TP63 gene have been previously described in AEC and RHS. All individuals included in this analysis were found to have mutations in either the SAM or TI domain of the TP63 gene. These mutations localize to exons 13 and 14 as previously reported in these conditions. Details of the molecular studies are provided by Rinne et al. 2009.


Our clinical evaluation recorded a number of abnormalities that were previously described in the literature. These included typical facial characteristics such as maxillary hypoplasia (88%), small mandible (65%), broad nasal bridge (76%), hypoplastic alae nasi (94%), thin vermillion border of the upper lip (88%) (Fig. 1), microstomia (88%), cleft palate (71%), and cleft lip (65%) (Fig. 2; Table I).

Figure 1.

Frontal and profile images of selected patients ages 4 months (upper left) to 30 years (lower right). Note variation in skin depigmentation in African-American individuals (upper and lower left) as well as hypoplastic alae nasi, flat midface, and pointed chin. The lateral portions of the eyebrows are often sparse (upper panel middle and 4th from the left). Scalp erosions are common and typically result in hair loss. Hair texture is also abnormal with a shiny “spun glass” appearance.

Figure 2.

Panel A: shows unrepaired bilateral cleft lip (arrow) and panel B shows a cleft hard palate (arrow) and trismus.

Table I. Summary of the Frequency of Phenotypic Features in 17 Individuals With AEC Syndrome
FeatureNumber (%)
History of ankyloblepharon12 (71)
Maxillary hypoplasia15 (88)
Small mandible11 (65)
Broad nasal bridge13 (76)
Hypoplastic alae nasi16 (94)
Thin vermillion border15 (88)
Short philtrum17 (100)
Trismus6 (35)
Cleft palate12 (71)
Cleft lip11 (65)
Hypospadias (males)7 (78)
2–3 toe syndactyly8 (47)
3–4 toe syndactyly3 (18)
3–4 finger syndactyly4 (24)
Split hand–foot malformation2 (12)

We performed two-tailed t-tests on a number of measurements and found significance for several clinical features including short philtrum length (mean −8.0 SD; P < .0001; 95% CI −9.0 to −7.1), small ears (mean −2.4 SD; P < .0001; 95% CI −3.3 to −1.5), decreased inner canthal distance (mean −1.1 SD; P < .01; 95% CI −1.9 to −0.4), and outer canthal distance (mean −2.1 SD; P < .0001; 95% CI −2.8 to −1.3). We also found significance for short stature (mean −0.8 SD; P < .01; 95% CI −1.3 to −0.2) and decreased weight (mean −0.7 SD; P < .05; 95% CI −1.3 to −0.1). Head circumference was not significantly different from the general population (mean −0.3 SD; 95% CI −1.1 to +0.57).

All subjects manifested hypotrichosis (sparseness of scalp and body hair) with thin hair, fine hair shaft caliber, and relatively light pigmentation when contrasted to the ethnic ancestry of the parents and siblings. Qualitative reduction in numbers of both eyelashes and brow hairs was also universal. Many subjects had mild hypotrichosis of the lateral third of the brow, but this could not be quantified in a rigorous manner. The results of the external and adnexal examination were generally normal, allowing for the expected thin skin and mild dysmorphism characteristic in the midface of persons with AEC syndrome.

Most evident was the absence of lacrimal puncta. Among all 17 subjects enrolled in this study and assuming two puncta (one upper and one lower) for each eye, the maximum number of expected puncta for this sample would be 68. However, the total number of puncta was 20, or a mean of 0.6 puncta per eye. The upper puncta were more commonly missing than the lower. Many subjects had symptomatic under production of tears and benefited from supplementation with artificial tears. Despite this, the ocular surface was generally normal, the corneal stroma normal, and the irides, anterior chamber structures, and the lenses normal and structurally sound. Dilated retinal examinations were not performed, since no unique or distinctive retinal pathology has been reported in the conditions of the AEC spectrum. No ankyloblepharon filiforme adnatum was seen in any of the patients examined, but 12 of 17 subjects reported this finding in the newborn period.

In our group of patients, 17 had limb abnormalities. The description of limb anomalies in the literature is limited to mild forms of soft tissue syndactyly and occasionally ectrodactyly. We, however, have found that in many cases the syndactyly observed was not just limited to partial syndactyly but rather complete (Fig. 3). The most common form of syndactyly seen involved 2/3 toes and was bilateral in all eight individuals (47%) with this finding. Three cases had 3–4 toe syndactyly that was noted to extend to the 5th toe in some cases (18%). In four cases, there was webbing and or syndactyly of the hands involving third and fourth digits (24%). Other limb anomalies included internal toe deviation (hammer toe deformities) in eight participants (47%). Many of those with internal toe deviation had very broad 1st toes. Camptodactyly limited to the hands was seen in two subjects. Cubital deviation of the digits was seen in two individuals. Generalized hypoplasia of the toes was also a common finding. Ectrodactyly was found in two subjects; limited to one limb in each case.

Figure 3.

Abnormalities of hands and feet ranged from very mild cutaneous syndactyly of one limb (A,B) and brachydactyly (C) to severe defects involving all four extremities. The feet in panels D and E are of the same individual, demonstrating syndactyly of digits 3–5 on the right foot (D) and 1–2 and 4–5 syndactyly on the left foot with split-foot deformity (E); his hands were normal. The foot and hand in panels F and G are of the same individual, demonstrating 2–3 and 4–5 syndactyly of the right foot (F) and bifid thumb with malformation of the index finger (G); all extremities in this individual had severe malformations.

From the nine male individuals examined in our group, seven had hypospadias (78%), and one had micropenis.

In our sample, two siblings who shared a frameshift mutation in the TI domain seemed to have severe scalp erosions; however, no other clear genotype–phenotype correlations were apparent in our cohort.


AEC syndrome, also known as Hay–Wells syndrome, has been characterized by a distinctive type of ectodermal dysplasia in association with orofacial clefting (cleft palate with or without cleft lip), craniofacial abnormalities, and eye findings including fusion of the eyelids and limb abnormalities. The diagnosis of AEC syndrome in our group is based on the clinical and molecular findings.

Due to the skin changes, many patients with AEC are often misdiagnosed with ichthyosis or epidermolysis bullosa (EB). While babies with AEC syndrome can have a collodian phenotype, which is often seen in many of the ichthyosis subtypes, they do not have the associated scaling, ectropion or eclabium. EB presents with non-inflammatory vesicles and bullae of the skin and mucosa that can lead to erosions, but the skin in EB infants is not typically erythematous; while the erosions of AEC syndrome are more superficial, more prominent on the scalp and associated with congenital erythroderma but not discrete blistering. As well, AEC syndrome is distinguished by the associated clefting and eye findings.

The differential diagnoses of AEC syndrome also includes other diseases caused by TP63 mutations such as EEC syndrome, ADULT syndrome, SHFM, and LMS. The limb abnormalities seen in our cohort of patients resemble those seen in SHFM or EEC, but with less obvious clefting of the hands and/or feet. Individuals with SHFM and EEC, however, do not have the extensive ectodermal involvement seen in AEC and RHS, including the scalp erosions, plantar and palmar involvement, as well as the severity of eye findings. EEC displays more severe limb abnormalities than those seen in AEC and RHS, with ectrodactyly being one of the hallmark findings of EEC. Our patients have subtler limb involvement and only two individuals presented with ectrodactyly limited to one limb in each case. In addition, the mutations for EEC syndrome are typically located at the 3′ end of the TP63 gene in the DNA binding domain contrary to the mutations in our cohort of patients. Even though LMS may display mutations in the SAM domain similar to AEC and RHS, it has distinct clinical features that include mammary gland aplasia/hypoplasia and lack of skin defects distinct to AEC and RHS.

With our systematic clinic evaluation, we were able to examine 17 patients with AEC syndrome, allowing us to give an estimate of the abnormalities present in this condition (Table I).

The main clinical features in AEC syndrome include skin findings secondary to the ectodermal dysplasia [see Julapalli et al., 2009]. These have been reported to include erythema and peeling skin at the time of birth, with erosions that may be limited or extensive. These findings may lead to infections and even sepsis and death. In addition, there may be paucity of the eccrine glands with associated hypohydrosis. Occasional skin abnormalities include palmoplantar keratoderma. The hair in these individuals is typically sparse, and their nails dystrophic or sometimes even absent, as often seen in other ectodermal dysplasias [Vanderhooft et al., 1993; Trueb et al., 1994; Rowan, 1996, see also Dishop et al., 2009]. The eyelids can be partially fused with fine strands of tissue joining the upper and lower lids, a phenomena known as ankyloblepharon filiforme adnatum; other eye findings include blepharitis, conjunctivitis, and absent lacrimal puncta [Hay and Wells, 1976; Weiss et al., 1992].

Other clinical findings include dental anomalies (conical and wide spaced teeth), hypoplasia of the maxillary region, broad nasal bridge, cup-shaped ears, ear canal atresia, short philtrum, small mouth, midface hypoplasia, and clefting of the palate, with or without cleft lip [see Cole et al., 2009; Farrington and Lausten, 2009]. There have been reports of patients with no cleft palate who present with velopharyngeal incompetence suggesting the presence of overt forms of clefting or submucous cleft palate and absent uvula [Greene et al., 1987]. Rare anomalies include supernumerary nipples (polythelia), limb abnormalities (syndactyly), short stature, heart defects, and genitalia anomalies.

We have also found a large percentage of genitourinary abnormalities in our male population. Genital abnormalities, including hypospadias, have been reported in RHS [Breslau-Siderius et al., 1991] and not as frequently in AEC syndrome. From the males examined, 78% had hypospadias. The hypospadias was mild in all cases and only involving the coronal region (grade I); one additional child was found to have a micropenis that increases the incidence of genitourinary abnormalities to 88% in our study population.

With regards to the limb anomalies, there has been no systematic documentation in the existing literature of AEC patients; in our sample, we have encountered a wide variety of limb anomalies including partial and/or total syndactyly, camptodactyly, and other digital abnormalities. We were surprised by the high incidence of limb anomalies in our group. In our review of the literature, we have not found similar descriptions. It is clear that the limb involvement in AEC is more significant than previously thought. This is not completely unfounded given mutations in the TP63 gene are involved in EEC, SHFM, and ADULT syndrome, which present with significant limb abnormalities. Interestingly, there have been ADULT syndrome mutations causing a frameshift localized to exons 13 and 14, in the SAM and TI domains where AEC mutations have been also found [van Bokhoven et al., 2001]. Patients with ADULT syndrome, however, do not share many of the other findings present in AEC, in particular the cutaneous findings.

Short stature and poor weight gain was also significant in our sample. These findings were associated with normal head circumferences. There have been anecdotal reports of short stature in AEC, but this growth failure has never been quantitated in a large cohort [see Motil and Fete, 2009].

In this largest population of AEC individuals ever examined to this date, we have been able to better delineate the frequency of known findings, as well as document other findings that have not been included or recognized as part of the phenotype. We have found that growth failure, genitourinary abnormalities (hypospadias in males), trismus, and limb abnormalities are commonly seen in AEC syndrome.

No clear correlation was found between the location (SAM or TI domain) of the mutation and the phenotypic features. The only correlation we have observed is in the case of two siblings who share a frameshift mutation in exon 14 in the TI domain. They presented with the most severe involvement of the skin seen in our cohort including extensive scalp erosions, severe alopecia, and blepharitis.

The TP63 gene represents a challenging example of a single gene with allelic heterogeneity and great variability. The TP63 related disorders are an example that defies the concept of lumping and splitting given the overlap of clinical features seen with different mutations and domains. Future functional studies and a better understanding of this protein will certainly allow us to better understand the genotype–phenotype given the complexity of mutation patterns.


This study was supported by the National Foundation for Ectodermal Dysplasias (NFED). We would like to thank Mary Fete for her organization of the symposium and also the patients and their families for participation in the clinical evaluations. Database entry and analysis was supported in part by the National Foundation for Ectodermal Dysplasias.