Article Text

Download PDFPDF

Further delineation of cardio-facio-cutaneous syndrome: clinical features of 38 individuals with proven mutations
  1. C M Armour,
  2. J E Allanson
  1. Department of Genetics, Children’s Hospital of Eastern Ontario, Ottawa, Ontario, Canada
  1. Dr C Armour, Department of Genetics, Children’s Hospital of Eastern Ontario, Ottawa, Ontario, Canada, K1H 8L1; carmour{at}cheo.on.ca

Abstract

Background: Cardio-facio-cutaneous syndrome (CFC) is a multiple congenital anomaly/mental retardation syndrome named because of a characteristic facies, cardiac anomalies, and ectodermal abnormalities. While considerable literature describes the main features, few studies have documented the frequencies of less common features allowing a greater appreciation of the full phenotype.

Methods: We have analysed clinical data on 38 individuals with CFC and a confirmed mutation in one of the genes known to cause the condition. We provide data on well-established features, and those that are less often described.

Results: Polyhydramnios (77%) and prematurity (49%) were common perinatal issues. 71% of individuals had a cardiac anomaly, the most common being pulmonary valve stenosis (42%), hypertrophic cardiomyopathy (39%), and atrial septal defect (28%). Hair anomalies were also typical: 92% had curly hair, 84% sparse hair, and 86% absent or sparse eyebrows. The most frequent cutaneous features were keratosis pilaris (73%), hyperkeratosis (61%) and nevi (76%). Significant and long lived gastrointestinal dysmotility (71%), seizures (49%), optic nerve hypoplasia (30%) and renal anomalies, chiefly hydronephrosis (20%), were among the less well known issues reported.

Conclusion: This study reports a broad range of clinical issues in a large cohort of individuals with molecular confirmation of CFC.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Named for the cardinal features, cardio-facio-cutaneous (CFC) syndrome is a sporadic multiple congenital anomalies/mental retardation condition with congenital heart defects, a characteristic facial appearance, and ectodermal abnormalities. Additional features include failure to thrive with severe feeding problems, moderate to severe developmental disability and short stature with relative macrocephaly. First described by Reynolds et al1 in eight patients, the syndrome has since been the subject of numerous reports. However, as for many rare syndromes, there remains insufficient information in the literature on the full breadth of clinical features.

CFC shows considerable phenotypic overlap with Noonan and Costello syndromes, making clinical diagnosis difficult, especially in the young child. Over the past few years, it has been determined that all three syndromes are caused by mutations in genes in the Ras-ERK signalling pathway; CFC by mutations in BRAF, MEK1, MEK2, and KRAS; Noonan syndrome by mutations in PTPN11, SOS1, KRAS, RAF1 and perhaps MEK1; and Costello syndrome by HRAS mutations.28 Molecular testing now yields definitive diagnosis in many borderline or questionable cases and allows analysis of homogenous groups of affected individuals. There remain approximately 10–40% of individuals with a clinical diagnosis of CFC in whom a mutation in one of the causative genes is not found.4 8 9

CFC International is a family support group that promotes research as part of its mandate. It supports a Biobank and has collected detailed medical information on registrants. In collaboration with CFC International, anonymised data were made available to us.

This clinical study of a cohort of 38 children and young adults with mutation proven CFC is the largest reported to date. While limited by the method of ascertainment (parental report and medical records that have not been systematically acquired), the data nonetheless provide new details of the breadth of phenotype and the frequency of particular features allowing improved information for parents and physicians responsible for clinical management. These can also be compared with the overlapping conditions, Noonan and Costello syndromes, to enhance our ability to distinguish between these three conditions.

METHODS

All individuals joining CFC International are asked to complete a questionnaire on their affected family member and to submit all medical records available plus a photograph. Information on study individuals was anonymised and submitted to the authors by CFC International. A total of 53 sets of data were collected. A database with all available data from each file was generated.

The majority of individuals underwent genetic testing for mutations in BRAF, MEK1, MEK2 and KRAS. Testing occurred at: Prevention Genetics, GeneDx, DuPont or the Rauen research laboratory (San Francisco Cancer Comprehensive Center). A total of 32 individuals had BRAF mutations, four had MEK1 and two had MEK2 mutations, bringing the total group of molecularly confirmed CFC patients to 38. Of the remainder, 12 did not undergo testing and three were tested and were negative. These 15 individuals were excluded from the analysis. It is unknown whether the three individuals who were negative had HRAS analysis. Twenty-one individuals have been previously reported,4 1012 the majority as part of the genetic elucidation of CFC.

Results are expressed as a percentage of those with a given feature out of those who had an informative answer (yes or no). Those who did not know the answer or for whom no data were available were not included in the denominator.

Ethics approval was obtained from the Children’s Hospital of Eastern Ontario Research Ethics Board.

RESULTS

Study population

There were 22 female (18 BRAF, 4 MEK1, 0 MEK2) and 16 male participants (14 BRAF, 0 MEK1 and 2 MEK2). The ages ranged from 13 months to 23 years, 4 months. There was no history of CFC in any family members.

The majority of parents of the affected individuals referred to their backgrounds as white (84%) while the remainder included black/white (1), Hispanic (1), Hispanic/white (3) and other (1).

Mean parental age varies year by year. The range in year of birth of this parental cohort is 1982 to 2005; median year of birth is 1998. The UK mean maternal and paternal ages for 1998 are 28.9 and 30.0 years, respectively (Office for National Statistics; http://www.statistics.gov.uk/STATBASE/ssdataset.asp?vlnk = 9526&More = Y). The US mean maternal age in 1998 was 27.1 years (http://www.cdc.gov/nchs/data/nvsr/nvsr51/nvsr51_01.pdf). Equivalent US paternal data are not available.

The average (SD) maternal age of the study population was 29.6 (6.1) years and the paternal age was 32.8 (7.5) years.

Intrauterine and neonatal periods

Polyhydramnios was a complication for 77% of respondents (19 BRAF, 3 MEK1, 1 MEK2). Prematurity (defined as birth before 37 weeks gestation) was also common and reported by 49% (14 BRAF, 1 MEK1, 1MEK2). A variety of other prenatal ultrasound findings were described including: hydronephrosis in two, one of whom also had bilateral pleural effusions, and single cases of enlarged ventricles, scoliosis, macrosomia, enlarged kidneys, increased nuchal thickness and borderline short femur.

Growth parameters

Of the 38 individuals, 27 (71%) had short stature (height below the 3rd centile) and 16 (42%) were below the 3rd centile for weight. All but one individual had birthweight >50% centile.

Of 33 sets of data on head circumference (measured at different ages ranging from birth to 15 years), four individuals (12%) were ⩾98% centile and hence had true macrocephaly. Only one of these four had hydrocephalus. Another 17 individuals (52%) had relative macrocephaly (defined by HC centile 1 SD greater than stature centile).

Two individuals were found to be growth hormone deficient while four others had normal growth hormone levels. Two different individuals were reported as having low or borderline low IGF-1. Thirty-three per cent of respondents had delayed bone age.

Cardiovascular

Heart disease is a cardinal feature of CFC and 71% of study individuals were affected (72% of BRAF; 75% of MEK1 and 50% of MEK2). The most common anomalies were pulmonary valve stenosis, hypertrophic cardiomyopathy and atrial septal defect/ventricular septal defect (ASD/VSD) (table 1). Nine individuals with BRAF mutations reported mitral valve disease including regurgitation, incompetence, stenosis and dysplasia. Aortic valve thickening and regurgitation were each reported in one child. Aortic coarctation was reported in two individuals. Subaortic stenosis was described in five individuals and resulted from a ridge, bulge, accessory valve tissue, an abnormal mitral valve or from hypertrophic outflow stenosis. There were single reports of tricuspid valve thickening, pulmonary artery branch stenosis, left superior vena cava to coronary sinus anomaly and a right ventricular outflow obstruction. No systematic information was collected on cardiac interventions and outcomes, but five individuals had had cardiac surgery and/or catheterisation. No deaths, cardiac or otherwise, were reported during the study.

Table 1 Frequency of some cardinal features among the 38 individuals

One individual (∼3%) had had supraventricular tachycardia and resting tachycardia.

Skin and hair

Sparse, curly hair with absent or sparse eyebrows (ulerythema ophryogenes) were among the most common hair findings (table 1). Eyelashes were described as thin in two individuals. Five individuals had other differences in the texture and quality of the hair (brittle, friable or coarse, wiry, frizzy, rough). Three individuals had receding hairlines (a 13-month-old female, a 7-year-old male, and a 23-year-old female who had it as a child). Six individuals had abnormal nails (hypoplastic or dysplastic).

The cardinal features of the condition, keratosis pilaris, hyperkeratosis, and ichthyosis, were commonly reported. However, nevi were the most frequently reported skin finding. Skin quality differences were noted including redundant skin (8), thick/doughy skin (3), rough/coarse skin (3), deep or unusual creases (2), and darker skin pigmentation than expected for the family (3). One child had prominent blood vessels and another had calloused hands and feet. Six individuals were reported to have eczema.

Neurology, development and behaviour

Neurological issues reported are found in table 2 and included hypotonia, seizures and hydrocephaly. Only one of the individuals with hydrocephaly required a ventriculo-peritoneal shunt. Five children had infantile spasms, and four had absence, generalised, or partial complex seizures. The rest were not well characterised. One individual had only one seizure and another’s presented for only a brief period following cardiac surgery. Other neurological issues included: ataxia (2), spasticity or increased tone (4), significant tongue thrusting (2), hemiparesis (2), cortical blindness (2) and deterioration in gait as a teenager (1).

Table 2 Neurological and behavioural issues

Brain imaging data, collected in a non-systematic fashion, was provided on 23 individuals. Numerous anatomical differences were reported including: hydrocephaly (2), ventriculomegaly or increased extra axial space (9), reduced white matter (6), thin corpus callosum (3), cerebral atrophy/small volume (3), delayed myelination (3), Chiari I malformation (1), arachnoid cyst (1), pachygyria (1), nodular heterotopia (1), migration abnormality (1) and cerebellar calcification (1). Twelve of the 23 reported individuals with abnormal anatomy had seizures. No imaging data were provided for the two individuals with hemiparesis or the sole individual with microcephaly.

The majority of parents described their children as having an engaging personality. Many of the other reported behaviours (table 2) are common in children with developmental delay or may be explained by other health issues—for example, irritability and chronic crying may be secondary to reflux. Many of these traits were reported to have improved over time.

Developmental data were available on 27 individuals over the age of 4 years. All had significant delay (table 3). Receptive communication was remarked to be better than expressive in nine individuals. The strongest academic skills achieved in this group were in a 23-year-old female (BRAF) who had late elementary school level abilities, and in a 7-year-old male (BRAF) who recognised two letters and was able to count to 20. Twelve individuals were reported to have used sign language and one used a picture exchange communication (PEC) system.

Table 3 Developmental data on 27 individuals over the age of 4 years at the time the questionnaire was completed (23 BRAF, 2 MEK1, 2 MEK2)

Musculoskeletal

Joint hyperextensibility and pectus excavatum/carinatum were the most common musculoskeletal features (table 4). Other reported features included: increased chest circumference (6), pes planus (5), crowded or overlapping toes (3), clinodactyly (3), wide anterior fontanelle (2), broad hands or feet (2), broad thumb/first toe (2), and syndactyly (2).

Table 4 Musculoskeletal, gastrointestinal and ophthalmic findings

Gastrointestinal

A significant number of individuals (92%) had gastrointestinal problems (table 4). When milder problems were eliminated, 71% of individuals had serious gastrointestinal symptoms, most as a consequence of dysmotility. Failure to thrive with profound feeding difficulty was present in 60%. Assisted feeding via nasogastric or gastrostomy tube was required. Many children were still gastrostomy tube fed in late childhood or early adolescence. Gastro-oesophageal reflux was reported by 55% (17 BRAF, 3 MEK1, 1 MEK2) of study individuals, seven of whom required fundoplication, while the remainder were treated with medication.

Two individuals required pyloroplasty. One individual had the procedure in conjunction with a repeat fundoplication due to continued reflux and vomiting; symptoms improved afterwards. Reason for pyloroplasty in the second is not stated. Three individuals had hepatomegaly, one of whom had a liver haemangioma. One case of liver haemangioendotheliomata without associated hepatomegaly was described. Anatomic differences included two individuals with an anteriorly placed anus, one with an anal tag and one with redundant perianal skin. There were only single cases of malrotation and splenomegaly.

Ophthalmology

The common ocular findings are found in table 4. Hypoplastic or dysplastic optic nerves were found in a significant proportion (42%) of individuals. Other eye related features included: anterior uveitis (1), cataracts (1), astigmatism (4), oculomotor apraxia (3), bluish sclera (2), and proptosis/exophthalmos (3).

Genitourinary

Cryptorchidism and kidney/bladder abnormalities were relatively common (table 5). Renal abnormalities included: dilated kidneys at birth, prenatal hydronephrosis, duplex collecting system (“extra outlet”), enlarged kidneys and vesicoureteric reflux (2). There were three reports of a large penis and one of a large clitoris. Nocturnal enuresis was described by three individuals; at 7, 11 and 18 years of age. There was inadequate information to comment on puberty.

Table 5 Other features reported

Hearing, ENT and oral

Table 5 lists the most common hearing and ear, nose and throat (ENT) issues. Eleven individuals underwent removal of the tonsils and adenoids. Two individuals were very sensitive to sound, and one had a cholesteatoma.

A variety of oral findings were reported including six individuals with dysplastic teeth (enamel hypoplasia, wide spacing, an extra lateral incisor and a double tooth). Six individuals exhibited drooling, and two had constant nasal drainage. Three individuals had differences in the quality of their voices (“hoarse/raspy/high pitched”). Anatomical differences included single cases of laryngomalacia, tracheomalacia, dacryocystocele, one individual who required pharyngoplasty, and two individuals with small nasal passages.

Sleep

Poor sleeping patterns were common among this group (table 5); however, there was one individual who was reported to sleep more than average. Sleep apnoeas were not uncommon and there was one report of hypopnoea. One individual with a BRAF mutation was described as having long laughing spells during sleep.

Respiratory and immune

Frequent respiratory infections, both upper and lower, were reported in nine individuals with BRAF mutations, and two individuals with MEK1 mutations. Chronic sinus infections were noted in three, and fungal infection of the skin in two individuals. One individual with a MEK1 mutation was found to have transient hypogammaglobulinaemia of infancy and 10 had allergies (drug, seasonal, food, etc).

Miscellaneous

One individual had von Willebrand disease. Two individuals had hypothyroidism that was described as mild, although both were treated. Age at diagnosis was not reported. One had precocious puberty and another individual had osteoporosis. Temperature homeostasis may be problematic as two individuals were documented as having high resting temperatures or pyrexia of unknown origin, and four were described as “sweaty” individuals.

The presence of prolonged feeding difficulties, seizures, central nervous system (CNS) abnormalities and optic nerve hypoplasia/dysplasia occurred in this study more frequently than expected. Only six individuals (16%) had none of the four features, 11 (29%) had one, 11(29%) had two, eight (21%) had three, and two (5%) had all four features. No imaging data were provided for the three mutation-negative individuals; however, none of the three remaining features were reported.

DISCUSSION

A number of the features described herein are in keeping with published data. Polyhydramnios1 12 and prematurity13 14 are known to be issues in the perinatal period. In this study both features are more common than previously reported. Data on growth parameters confirm what is already described—most have a birth weight greater than the 50% centile, with 60% developing failure to thrive. More than two thirds have short stature and just over half have macrocephaly (both absolute and relative). These features are also typical of Costello and Noonan syndromes.

The proportion of individuals with specific cardiovascular anomalies is comparable to that in recent publications.9 12 15 16 The main abnormalities—pulmonary valve stenosis, atrial septal defect and hypertrophic cardiomyopathy—are seen in all three syndromes. Pulmonary valve stenosis is more common in Noonan syndrome and less so in Costello syndrome, while hypertrophic cardiomyopathy is more common in Costello syndrome and less common in Noonan.16 The combination of pulmonary stenosis and ASD has been described as typical of CFC syndrome,12 15 although there were only four individuals with this combination. Arrhythmias, specifically atrial tachycardias, are more characteristic of Costello syndrome and reported to occur in approximately 33%.17 They are much less common in CFC as reported here (one individual (∼3%) with supraventricular tachycardia) and elsewhere (8%).15 Arrhythmia is the main cardiovascular feature that can help differentiate CFC, Costello and Noonan syndromes.

Skin findings are key features of the syndrome and ichthyosis, hyperkeratosis and keratosis pilaris are more characteristic of CFC than of Costello or Noonan syndromes, and may help to distinguish them. However, facial keratosis pilaris and curly hair are prevalent in individuals with Noonan syndrome who have SOS1 mutations.7 Nevi have been described in CFC,13 and the frequency documented here is high. However, this is not surprising in light of the finding that BRAF is mutated in 82% of nevi.18 Differences in hair texture and distribution are frequent. While curly hair can be seen in both Costello and Noonan syndromes, ulerythema ophryogenes is more characteristic of CFC.

There have been rare reports of persons with CFC who had normal development. We would question the diagnosis if intelligence is normal.19 Nava et al recently reported two individuals with MEK1 mutations, one with borderline and one with normal mental development.8 The phenotype in these two children was said to resemble Noonan syndrome, not CFC. Developmental delay and moderate/severe mental retardation were seen in 100% of individuals over 4 years of age in our cohort, higher than reported elsewhere.12 This may reflect the fact that families with more severely affected children are more motivated to join a support group. While all motor skills are delayed, the majority learns to walk (if only with a walker). Expressive communication is quite impaired as less than half used two or more words together. Receptive communication, although not specifically addressed, was felt by many parents to be significantly better than expressive language. In comparison, individuals with Noonan syndrome have normal IQ with one third requiring special education. Costello syndrome has less severe mental impairment than CFC.20 21 Description of the personality of CFC is similar to the warm and friendly personality described in Costello syndrome.21 Although occasionally noted, the anxiety that is described in Costello syndrome was not pervasive in these individuals with CFC.

Seizures are clearly much more common than has been appreciated (49%). Earlier reports have not emphasised this problem. Two recent reports described seizures in 3/13 (23%) and 9/25 (36%) of patients.9 15 In Costello syndrome seizures are present in approximately 20%.22 CNS abnormalities are not frequently described in CFC, although Zenker et al reported three individuals with CFC due to KRAS mutations who had cerebral abnormalities.23 Gripp et al reported four of 13 patients with CNS abnormalities, including hypoplastic corpus callosum, increased extra-axial fluid and a patient requiring a ventriculo-peritoneal shunt.15 Our study suggests that individuals with BRAF, MEK1 and MEK2 mutations may also have CNS abnormalities more frequently than previously recognised. We hesitate to draw too many conclusions regarding these data given that imaging records were not collected in a systematic fashion, and many families did not submit any imaging descriptions.

The presence of gastrointestinal issues in CFC has been recognised and postulated to reflect a more severe phenotype11; however, only recently has it been commented upon in more detail.12 24 In this study, constipation, swallowing difficulties and gastro-oesophageal reflux were problematic for two thirds, and more than half of the children required medical intervention in the form of a nasogastric or gastrostomy tube, frequently into late childhood or early teenage years. Similar issues are described in Costello21 and Noonan syndromes, though the duration is typically much shorter; in Costello syndrome the feeding difficulties often resolve by 2–4 years and in Noonan syndrome by 18 months20 21 Although reported, hepatomegaly, splenomegaly and malrotation were not common, in contrast to the original cohort described by Reynolds et al in which four of the eight patients had splenomegaly.1

This study highlights many of the musculoskeletal features that have been described.13 Similar musculoskeletal findings such as pectus excavatum, kyphosis and scoliosis are described in Costello and Noonan syndromes.20 21 Hyperextensibility of the small joints of the hands is notable with ulnar deviation producing a characteristic wrist drop in Costello syndrome.25

Ophthalmological abnormalities such as esotropia and nystagmus have been recognised since the publication of Reynolds et al, yet few publications have addressed the ocular phenotype.1 12 26 This study reveals a relatively high frequency of eye issues. Forty-two per cent of individuals have dysplastic or hypoplastic optic nerves; an unusual finding not reported in Costello syndrome and rare in Noonan syndrome, thus a useful discriminator.

Cryptorchidism was noted in one of the original four male patients described.1 It is seen in 10% of individuals with Costello syndrome,27 and is even more common (60–80%) in Noonan syndrome.20 The renal abnormalities, albeit mild, are similar to those seen in Noonan syndrome. The possibility of large genitalia as a feature exists, but will require further study to clarify.

Chronic otitis media was reported in many individuals (55%), most requiring placement of multiple sets of tympanostomy tubes. Frequent upper respiratory infections may be contributory.

Unlike Noonan syndrome, there does not seem to be a predilection for bleeding diathesis, with a single individual reported to have von Willebrand disease.

A major and important difference between the three conditions is the absence of neoplasms in this cohort, and in CFC in general. There have been case reports of malignancies in CFC, with one report of hepatoblastoma in a post-heart transplant patient with a MEK1 mutation and two cases of acute lymphoblastic leukaemia (one with a BRAF mutation and the other reported before gene discovery).2830 An individual with CFC has been reported to have rhabdomyosarcoma, although it is likely this individual has Costello syndrome.31 Individuals with Costello syndrome are at increased risk of developing a malignancy, with approximately 17% developing a neoplasm such as rhabdomyosarcoma, neuroblastoma or transitional cell carcinoma of the bladder.32 Individuals with Noonan syndrome and PTPN11 mutations are at a greater risk of juvenile myelomonocytic leukaemia, and there have also been reports of acute myelogenous and lymphatic leukaemia, neuroblastoma and rhabdomyosarcoma.20

Many of the clinical features described herein are in keeping with recently described series8 9 15 and also in keeping with the CFC index proposed by Kavamura et al.10 Cardiac abnormalities were seen with similar frequency: 71% of our study group, versus 77% of those in Nava et al, 62% in Gripp et al and 84% in Narumi et al. Mental retardation was universal in our study, and in those of Nava et al and Narumi et al. Hair and skin features were comparable between the studies. It is difficult to compare skin findings between studies since the categories are presented differently. The pattern of normal/large birthweight with postnatal development of short stature is also consistent, though proportions vary slightly. Some differences between series include hypotonia (94% this study vs 56% Narumi et al, 68% Nava et al and 77% in Gripp et al), seizures (49% this study vs 16% in Nava et al, 23% in Gripp et al and 36% Narumi et al), failure to thrive (67% this study vs 81% Nava et al and 100% Gripp et al), and polyhydramnios (77% this study vs 46% Gripp et al and 54% in Nava et al). The most important difference is the high likelihood of significant gastrointestinal dysmotility and optic nerve hypoplasia in our study, which are not described in other series. The differences between series are likely related in part to methods of ascertainment of patients.

This study clearly has limitations. Some data are derived from parental questionnaire and hence may suffer from recall bias. Medical records were not systematically collected. Some children have seen paediatric subspecialists while others have not. The study also reflects a subset of individuals who have joined CFC International. Parents of children with greater needs may be more inclined to seek out such resources. With recent availability of molecular testing, children and adults with milder features will come to attention and the data will evolve.

This study summarises the most frequent medical issues in 38 individuals with CFC syndrome. Knowledge of the causative mutation allows confidence in the diagnosis and resulting data. We have been able to confirm and, in some instances, refine information on cardinal features already in the literature, and provide new data on less frequent manifestations. We want to emphasise the importance of prolonged feeding difficulties, central nervous systems anomalies, seizures, and optic nerve hypoplasia as diagnostic clues since 84% had at least one of these features, and 5% had all four. We suggest that these are key features that may help to distinguish CFC from Noonan and Costello syndromes.

Acknowledgments

We are greatly indebted to the patients and their families for their willingness to share their personal information and to CFC International, particularly Brenda Conger, for the considerable efforts in collecting and organising the data.

REFERENCES

Footnotes

  • Competing interests: None declared.

  • Ethics approval: Ethics approval was obtained from the Children’s Hospital of Eastern Ontario Research Ethics Board.

  • Patient consent: Informed consent was obtained for the publication of the persons’ details in this report.