MATERIALS AND METHODS

Eighty-eight cases of FS satisfying the published diagnostic criteria were ascertained using the search terms “Fraser syndrome” or “cryptophthalmos” on the OMIM database (www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?db=OMIM) and PubMed (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi).^3–5,7–54

Twenty-nine cases of CO or eyelid colobomas did not satisfy the diagnostic criteria for FS and were obtained using the same methods.^{39,52,55–69} We did not include the cases reported in Philip et al⁷⁰ and King et al⁷¹ as insufficient clinical details were provided. We also noted that the details of the four cases reported by Vanlieferinghen et al⁷² were the same as described in Francannet et al¹⁹ and we have only included these cases once as Francannet et al.¹⁹ The patients reported by Pe’er and BenEzra⁷³ and Pe’er et al¹¹ are also the same and the clinical details in the latter paper were used. Similarly, the patient reported by Chattopadhyay et al⁷⁴ appears to be the same as the patient in Jagtap et al³⁷ and has been included once as Jagtap et al.³⁷ We included the cases of Feldman et al,⁵⁵ Ohtsuka et al,⁵⁶ and Wiznitzer et al⁷ as these cases were published in close chronology to and omitted from the review of Thomas et al.²

For each case, the following information was collated when available: parental ages, consanguinity, family history, sex of proband, karyotype, and details of pregnancy including length of gestation, weight, and other measurements at the time of birth. We recorded eye malformations, airway malformations, digital abnormalities, renal malformations, genital abnormalities, cardiac malformations, gastrointestinal malformations, cerebral malformations, orofacial clefting, skeletal defects, abnormalities of the thymus, developmental status, and other phenotypic findings.

RESULTS

Data from the 117 cases are shown in tables 2-18. There were 57 females (48.7%), 54 males (46.2%), and six (5.1%) in whom the sex was not able to be determined. Eighty-eight patients (75.2%) satisfied the published diagnostic criteria for FS, whereas 29/117 (24.8%) did not (see Materials and methods for listing of individual cases). Consanguinity was present in 29 cases (24.8%, data not shown) and the most common consanguineous union was first cousins.^{7,11,13,19,25,29,42,50,55} Of those born to non-consanguineous parents, one child was from gypsy parents³³ and one child was born to parents from the same village.⁹ Forty-eight patients (41%) had a significant family history of a relative with CO or physical findings suggestive of FS (data not shown). The oldest subjects were alive in the fourth decade of life.⁵⁸

The average paternal age (rounded to the nearest whole year) was 27 years of age and the average maternal age (rounded to the nearest whole year) was 24 years of age (data not shown). Thirty-six patients were reported to have a normal female karyotype and 21 patients had a normal male karyotype (data not shown). Two patients had an inversion of chromosome 9 that was considered to be unrelated to their physical findings ((46,XX,inv(9)(p11q21) and 46,XY,inv(9)(p11q21)).²¹

Pregnancy

Oligohydramnios was the most frequent complication during pregnancy (17.1%, table 2). The majority of babies in whom gestational age was stated were born at term (data not shown).

CO and ocular malformations

CO was present in 103/117 (88%) of cases and was bilateral in 62 cases (53%) and unilateral in 32 cases (27.4%, table 3). The type of CO was often not provided but complete CO appeared to be the most common form (data not shown). CO was commonly associated with a tongue of hair extending across the lateral face (40/117, 34.2%, table 4), absent eyebrows or eyelashes (34/117, 29.1%), and coloboma of the eyelid (21/117, 17.9%). Other abnormalities included microphthalmia (25/117, 21.4%), anophthalmia (7/117, 6.0%), and corneal opacification (12/117, 10.3%).

Digital anomalies

Syndactyly was the most common digital abnormality (72/117, 61.5%, table 5). Syndactyly of the hands and feet was present in half of the cases with syndactyly. Brachydactyly, nail hypoplasia, and abnormal palmar creases were present in less than 10% of patients.

Genital malformations

Ambiguous genitalia were found in 20/117 (17.1%, table 6). In females, clitoromegaly was the commonest genital abnormality (21/57, 36.8%). Bicornuate uterus, uterine hypoplasia, vaginal agenesis, and synechiae or hypoplasia of the labia were present in more than 8% of females. In males, cryptorchidism (17/54, 31.5%), micropenis, phimosis, chordee, hypospadias, and scrotal hypoplasia were noted.

Nasal malformations

Nasal anomalies were common with 24 having a broad nose or nasal bridge (20.5%, table 7), 13 with a depressed or flat nasal bridge (11.1%), and 18 with a bifid nose or a midline nasal groove (15.4%). Coloboma of the nares or notched nares were present in 13/117 (11.1%).

Malformations of the ears

Malformed and/or low set ears,(63/117, 53.8%, table 8), microtia (19/117, 16.2%), and atresia or stenosis of the external auditory meatus (21/117, 17.9%) were recorded.

Malformations of the airway and lungs

Laryngeal stenosis or atresia was reported in 36/117 (30.8%, table 9). Choanal stenosis or atresia (7/117, 6%) and subglottic stenosis (10/117, 8.5%) were also described.

Orofacial clefting

Clefting of the lip, palate, uvula, or upper gum or a combination of clefts were noted in 13/117 (11.1%, table 10). A high arched palate was found in 14/117 (12%).

Musculoskeletal anomalies

Absence or hypoplasia of the orbital or skull bones (12/117, 10.2%) and defects in skull ossification (8/117, 6.8%) were among the most frequent musculoskeletal anomalies (table 11). Talipes (10/117, 8.5%) and abnormalities involving the pubic symphysis (9/117, 7.7%) were relatively common.

Gastrointestinal malformations

The commonest malformation was imperforate anus (15/117, 12.8%, table 12), but anal atresia and anal stenosis were mentioned in 9/117 (7.7%) and 8/117 (6.8%) cases, respectively. Thirteen patients had a low set umbilicus (11.1%).

Renal malformations

Bilateral renal agenesis with or without agenesis of the ureters was present in 27 of patients (23.1%, table 13) and unilateral renal agenesis with or without ureteral agenesis in 26 cases (22.2%). The bladder was small or absent in 20 (17.1%) and cystic dysplasia of the kidneys was reported in 14 patients (12%).

Cerebral malformations

Hydrocephalus was recorded four times^3,4,31,55 and polymicrogyria or abnormal brain gyri were seen in three cases.⁴ Two patients had encephaloceles.^39,65 Other single findings were mild cerebellar hypoplasia,¹⁶ holoprosencephaly with hydromelia of the spinal cord,³ periventricular leucomalacia,¹² diffuse gliosis of the brain,⁴¹ and low brain weight.²⁶

Cardiac malformations

Cardiac malformations included hypertrophy of the left ventricle,³³ hypertrophic heart,¹⁶ a variant of Ebstein anomaly,³⁶ coarctation of the aorta,⁴ an atrial septal defect,³⁷ an interventricular communication,¹⁵ and a truncus arteriosus and a ventricular septal defect.⁴ One patient had complex heart disease with dysplasia of the pulmonary and aortic valves and endocardial fibrosis.⁵⁰ A patent foramen ovale and patent ductus arteriosus were present in three cases^26,40,44 and one patient had a patent ductus arteriosus and dilated coronary sinus.²⁶ Dextrocardia³ and transposition of the great vessels⁵⁵ were also noted.

Thymic abnormalities

Absence or hypoplasia of the thymus^3,36 and two thymuses²⁶ were described.

Developmental delay/psychomotor retardation

Information regarding intellectual development was provided in few cases. Developmental delay was present in five patients^{25,37,49,52,61} and delayed motor development was found in two.^3,13 Speech delay was specified in one subject.¹⁴ Development was normal in 13 cases.^{3,58,60,62,65–67,69} Hypotonia was reported twice.^13,25

Dysmorphological findings

There was no recognisable facial phenotype (table 14). Twenty-five (21.4%) subjects had hypertelorism or pseudohypertelorism, 11/117 (9.4%) had micrognathia, 8/117 (6.8%) had microstomia, and 8/117 (6.8%) had a short neck.

Fetal and postnatal growth

In babies born at or after 33 weeks of gestation, the majority had normal growth parameters (data not shown). Nine out of 50 (18%) had a growth parameter below the 3rd centile (data not shown). Fraser syndrome is compatible with normal postnatal growth but both microcephaly³⁷ and macrocephaly⁶² have been recorded and one patient had a final height of 128 cm (<3rd centile, 50th centile for age 8-9 years), whereas another case had growth hormone deficiency.²⁵

Survival

The age at reporting or age of death are shown in table 15. In those who died in the first week of life, the commonest causes of death were laryngeal stenosis/atresia or respiratory insufficiency,^4,16,26,32 obstructive uropathy or bilateral renal agenesis, ^{4,7,16,21,31,34,41} or a combination of laryngeal and renal malformations.^12,40,46 Pulmonary agenesis and bilateral renal agenesis were seen in one infant.¹¹ Patients who were alive at 10 years of age or older had fewer major malformations (data not shown) and only one of the 10 satisfied the diagnostic criteria for Fraser syndrome. Interestingly, there was a strong phenotypic similarity and concordance of the degree of severity of the disease in families for both severely affected^{4,19,21,32,36,44,50} and mildly affected^62,66 sibs. However, intrafamilial variation was seen in the family described by Hancheng,⁵⁸ in which two sibs survived to the fourth decade of life and one died at 2 months of age.

DISCUSSION

We have compiled the phenotypic features of 117 patients with CO and FS (figs 1 and 2) reported since the review of Thomas et al² in 1986 and compared the incidences of phenotypic findings with previous reviews (table 16).^2,3 Our aim was to review the phenotypic manifestations associated with CO and FS and to examine the efficacy of the published diagnostic criteria.² The frequency of malformations in this patient group does not differ significantly from the frequencies reported by Gattuso et al.³ However, they are lower than those reported by Thomas et al² owing to selectivity of the latter paper in including only patients in whom the presence or absence of a feature had been documented. The incidence of published malformations could also be skewed because of the preferential inclusion of rare features and complications or severely affected patients in medical publications in a well described syndrome.

Our data show an increased incidence of consanguinity (24.8%) compared to the incidence of 15% published by Thomas et al,² consistent with autosomal recessive inheritance. The prevalence of FS has previously been estimated to be approximately 11 cases in 100 000 live births.⁴⁵

CO is considered to be the single most important diagnostic malformation in FS. Complete CO is usually bilateral and can be associated with absence or poor development of the eyebrows, eyelashes, gland structures and conjunctival sac, microphthalmia, symblepharon, and abnormalities of the anterior chamber of the eye.^8,75 In incomplete or atypical CO, rudimentary lid structures with small, lateral conjunctival sacs are present with small palpebral fissures, microphthalmia, and symblepharon.^8,75 CO should be differentiated from microblepharon (vertical shortening of the eyelids)⁷⁶ and mesodermal corneal metaplasia.^77,78

Abortive CO or congenital symblepharon and ablepharon are descriptive terms used to describe an upper eyelid without a well defined margin that is adherent to the cornea, often with a small globe and keratinisation of the cornea.⁸ Symblepharon and ablepharon have been considered by some authors to be an abortive form of CO^8,79 and by others to constitute a separate pathological entity.⁸⁰ The degree of CO and the range of ocular abnormalities in FS was very variable (tables 3 and 4) and many patients had complete CO in one eye with incomplete or abortive CO affecting the other eye (data not shown).

We found patients who had symblepharon and ablepharon and phenotypic features consistent with FS, suggesting that symblepharon is part of the ocular manifestations of FS. The ocular abnormalities in FS were almost all confined to the anterior chamber of the eye with the exception of six patients who had hypoplasia or atrophy of the optic nerve (table 4).

In our series, 14 (12%) patients did not have CO.^{4,9,16,33,34,41,45,50,53,55,62} Most of these patients had other ocular abnormalities consistent with FS (for example, corneal clouding,⁴ sclerocornea,³³ microphthalmia,⁵⁵ anophthalmia,^{34 62} microcornea,⁴¹ and a lateral tongue of hair,⁶² and 78.6% satisfied the diagnostic criteria for FS (data not shown). However, in the absence of CO, anterior chamber abnormalities could conceivably be confused with other malformation syndromes (for example, Walker-Warburg syndrome or Peters’ plus syndrome). The presence of microphthalmia or anophthalmia without CO in a sporadic case constitutes a further diagnostic dilemma. For example, a male reported by Glanz et al⁸¹ as having Lenz microphthalmia because of short palpebral fissures could be considered to satisfy the diagnostic criteria for FS with syndactyly, cryptorchidism and hypospadias, renal hypoplasia, and a cleft palate.

The diagnosis of FS must therefore be made with caution in patients who do not have CO or a family member with CO, even if other ocular abnormalities consistent with FS are present. For example, in the patient reported by Martinez-Frias et al³³ who had sclerocornea, syndactyly, ambiguous genitalia, a furrowed nasal tip, low set, posteriorly rotated ears, and an umbilical hernia, the authors did not consider that the diagnosis of FS was correct although this patient did fulfil the diagnostic criteria. Without CO, FS can be overdiagnosed because of the relatively high frequency of digital and genital abnormalities in multiple congenital anomaly syndromes.

In the patients with CO who did not satisfy the diagnostic criteria for FS, only one had CO without any other phenotypic findings and with an unremarkable family history.⁶⁰ Autosomal recessive inheritance in isolated CO should still be considered and three affected patients with CO as the sole physical finding were born to consanguineous parents.³⁹ A common anomaly found in addition to CO without FS was a tongue of hair extending from the scalp to the lateral eyebrow (table 4).⁵⁸ We consider that this finding makes an underlying diagnosis of FS with the implication of autosomal recessive inheritance more likely and would consider it as having at least equal importance to a minor diagnostic feature.

The incidence of syndactyly in this patient cohort was 61.5% (table 5), less than the frequency of 79% reported by Thomas et al.² However, the syndactyly in many patients was distinctive because of the involvement of both upper and lower limbs (table 5) and the extensive nature of the cutaneous webbing, which frequently included all digits (data not shown). A range of external and internal malformations of the genitalia were described, fully justifying the inclusion of these abnormalities as major diagnostic criteria.²

The minor criteria of malformations of the ears and nose can be non-specific and more weighting should be given to these features if they are included in the distinctive anomalies found in FS, such as coloboma of the nares or notched nares (table 7). Orofacial clefting (table 10) and mental retardation were infrequent and these criteria were rarely helpful in establishing the diagnosis. No mention has been made of gastrointestinal tract malformations in the diagnostic criteria and we would recommend consideration of malformations such as anal atresia, rectal atresia, anal stenosis, and imperforate anus as minor criteria owing to the occurrence of these features in more than 25% of affected subjects (table 12). A low set umbilicus may be more frequent and hence more diagnostically useful than an umbilical hernia (table 12). Renal agenesis has previously been considered to be an important diagnostic feature in FS and we would agree that this abnormality is pertinent to the diagnosis.^82–84

The FS phenotype is complex and pleiotropic and therefore has significant overlap with other malformation syndromes. We were interested to determine if there were distinct patterns of physical features within the FS phenotype. Recently, more complex modes of inheritance involving modifier genes or three altered alleles have been described for a different autosomal recessive condition, Bardet-Biedl syndrome (BBS).^85,86 We therefore speculated that the more complex modes of inheritance identified in BBS could also be described for other pleiotropic syndromes and that, in some cases, the phenotypic consequences of the different genes or pathways may be identifiable within a syndromal phenotype.

No phenotype-genotype correlation has as yet been consistently described for BBS patients. However, syndactyly, imperforate anus, Hirschsprung disease, cardiac defects, and female upper genitourinary tract malformations have previously been found with increased frequency in BBS patients with hydrometrocolpos owing to vaginal agenesis compared to unselected BBS patients (data not shown). In this patient cohort, we found seven females who fulfilled the diagnostic criteria for FS and who had vaginal agenesis or vaginal atresia. It was surprising to find that all of these patients also had at least one other finding consistent with the above pattern such as anal abnormalities, renal malformations including renal agenesis or renal cysts, and bicornuate uterus (table 17). In addition, cardiac abnormalities were identified in two patients^41,50 and one patient did not have CO.¹⁶ The finding is more striking if one considers the relative rarity of some of these findings in FS (table 18). This phenotypic subset has some similarity to the MURCS association (Muellerian duct aplasia, renal aplasia, and cervical dysplasia) and the Rokitansky malformation sequence of vaginal atresia and uterine hypoplasia or a bicornuate uterus and renal agenesis.^87,88 Interestingly, corneal anaesthesia and punctate epithelial opacities have been described in the MURCS association⁸⁹ and a child with bilateral microtia and hypoplasia of the external ear canals, a cleft palate, hypoplastic thumbs, renal agenesis, pulmonary agenesis, and genital hypoplasia has been considered to have physical features consistent with MURCS association and Nager acrofacial dysostosis.⁹⁰

Similarly, we found that laryngeal stenosis was present in 35/117 of patients with FS (29.9%) and stenosis of the external auditory meatus in 21/117 of FS patients (17.9%). Both of these malformations were present in 11 patients (52.4%, tables 12 and 13). This association is significant with a p value of 0.046. Although hypertelorism, hypopspadias, and laryngeal malformations are found in both Opitz syndrome and FS, there did not appear to be any association of these features in this patient group (data not shown).

Modifier genes are important determinants of phenotypic variation and have been shown to be clinically important in diverse conditions, including sensorineural deafness,⁹⁰ cystic fibrosis,⁹¹ hypertrophic cardiomyopathy,⁹² early onset glaucoma,⁹³ and keratin filament disorders.⁹⁴ Modifier genes can also have tissue specific effects.⁹⁵ The significance of modifier genes in the generation of the variability of the FS phenotype is unknown but is not supported by the strong familial concordance in phenotype in many reported cases (see above).

We therefore suggest that subsets of physical anomalies or phenotypic modules can be conserved across different syndromes and that they may prove to be a useful means for the delineation of specific abnormalities within a syndrome and for the determination of relevant molecular screening tests. The pathogenesis of phenotypic modules could include disruption to a morphogenetic field or a developmental field,⁹⁶ mutation specific effects, or malfunction of temporally distinct genes. Consideration of the physical findings in a syndrome as a series of interacting phenotypic modules may also be a useful method for determining phenotype-genotype correlations in the future.

CONCLUSION

We have reviewed 117 patients diagnosed with FS and CO since the publication of the diagnostic criteria for FS by Thomas et al.² The diagnosis should be made with caution in probands and families without CO, although in the presence of typical findings, CO is not essential for the diagnosis. The physical features of orofacial clefting, umbilical hernia, and mental retardation were less useful in making the diagnosis, whereas gastrointestinal tract malformations may be helpful.

We also found that patients with vaginal agenesis and FS had a pattern of additional malformations previously described in MURCS association and BBS. This suggests that there is conservation of a subset of phenotypic features between different syndromes and that unusual mechanisms of inheritance such as modifier genes or triallelic inheritance may be present in malformation syndromes other than BBS.