Silver–Russell syndrome (SRS) is characterised by intrauterine growth restriction, poor postnatal growth, relative macrocephaly, triangular face, asymmetry and feeding difficulties. As many of these features are non-specific, clinical diagnosis of SRS remains difficult. Hypomethylation of the imprinting control region (ICR) 1 on chromosome 11p15 and maternal uniparental disomy (mUPD) for chromosome 7 are found in up to 60% and around 5–10% of patients with SRS, respectively. Patients with ICR1 hypomethylation are more likely to have classical features of SRS, including asymmetry; patients with mUPD7 are more likely to have learning difficulties, particularly speech problems, although these are usually mild. As features vary widely in severity, clinicians should have a low threshold for genetic investigation of patients with features suggestive of SRS.
Statistics from Altmetric.com
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.
In 1953 Silver et al1 described two children with low birth weight, asymmetry and growth restriction. The following year, Russell2 described five children with similar features, two with asymmetry. Since 1970, the term Silver–Russell syndrome (SRS) has been used to describe individuals with the characteristic features of intrauterine growth restriction (IUGR), poor postnatal growth, relative macrocephaly, a triangular facial appearance, fifth finger clinodactyly and body asymmetry.3 4 Feeding difficulties are also seen the majority of affected children.
The frequency of SRS is not easily determined as the features are non-specific and vary widely in severity. In addition, signs are most striking in infancy and early childhood, making assessment of older children difficult. It is rarely easy to make a confident clinical diagnosis and it is likely that this condition is both over- and underdiagnosed.
Within the last few years it has become possible to detect molecular abnormalities in around half of the patients with a clinical diagnosis of SRS. Within this group, two main molecular mechanisms are known to be involved: maternal uniparental disomy for chromosome 7 (mUPD7) and methylation abnormalities of chromosome 11p15. However, there remain a significant number of patients with clinical features suggestive of SRS in whom the molecular cause is currently unknown.
The first molecular abnormality to be identified in a significant proportion of patients was mUPD7 (where both copies of chromosome 7 are inherited from the mother and none from the father).5 Evidence from a number of studies suggests that mUPD7 can be identified in around 5–10% of cases.6 It is likely that the growth failure associated with mUPD7 arises from altered expression of an imprinted gene(s) on chromosome 7. This could be either over-expression of a maternally expressed growth suppressor or underexpression of a paternally expressed growth promoter. However, although several imprinted genes on chromosome 7 have been identified, none has been directly implicated in SRS to date.
More recently, attention has focused on chromosome 11p15, which contains a cluster of imprinted genes crucial to the control of fetal growth. The region consists of two imprinting centres (ICRs) (figure 1): the telomeric ICR1 regulates expression of IGF2 and H19 and the centromeric ICR2 controls expression of CDKN1C, LIT1 (KCNQ10T1) and other genes. Disturbances of this region were already known to result in the overgrowth disorder Beckwith–Wiedemann syndrome (BWS).7 In 2005, Gicquel et al demonstrated loss of paternal methylation at the H19-IGF2 ICR1 in five out of nine patients with typical features of SRS.8 This defect was associated with downregulation of expression of the growth promoter IGF2. ICR1 hypomethylation is the major abnormality found in SRS patients within this region and can be detected in up to 60% of SRS patients.6 To date, only one patient has been reported with a maternally derived duplication restricted to ICR2.9
The mechanism of growth restriction in ICR1 hypomethylation is still poorly understood. Although reduced IGF2 expression has been demonstrated in fibroblast culture,8 patients with ICR1 hypomethylation have normal serum IGF-II levels.11 IGF-II has a central role in fetal growth, being predominantly paternally expressed in the placenta/antenatally. After birth, serum IGF-II is mainly produced in the liver where IGF2 imprinting is relaxed. Reduced IGF2 expression with ICR1 hypomethylation may therefore have its main effect on prenatal growth. IGF-II may also remain low in the tissues postnatally despite normal serum IGF-II levels.
The frequency of clinical features in SRS patients with confirmed ICR1 hypomethylation and mUPD7 are compared in table 1.
Initial reports of patients with mUPD7 suggested that they have a milder phenotype.10 Severe feeding difficulties, speech delay and excessive sweating were still common, but typical facial features of SRS and asymmetry were observed less frequently. The majority of patients with a confirmed genetic diagnosis of ICR1 hypomethylation appear to have a typical SRS phenotype with IUGR, postnatal short stature, characteristic craniofacial features and, in many, asymmetry.11,–,17 Studies comparing clinical features in patients with either ICR1 hypomethylation or mUPD7 have described a less severe or ‘classical’ phenotype in mUPD7 patients.12 14,–,16 However, it is not possible to clearly differentiate the two groups on clinical grounds alone.14 16
The majority of studies have found no evidence for ICR1 hypomethylation in cohorts of patients with isolated pre- or postnatal growth restriction.12 13 16 18 However, there are a small number of reported ICR1 hypomethylation patients presenting primarily with hemihypotrophy or with milder degrees of pre- and/or postnatal growth failure.11 17 19 20
Some reports have suggested that the severity of the SRS phenotype may reflect the level of ICR1 hypomethylation.8,11 13 19 However, a recent study found no evidence of correlation between the level of ICR1 hypomethylation and a number of clinical variables including birth weight, postnatal height, feeding difficulties, developmental delay, asymmetry and/or congenital anomalies.14 Alternative explanations for the wide range in clinical severity seen in association with ICR1 hypomethylation include tissue level mosaicism, selective hypomethylation of H19 or IGF2 and/or the presence of hypomethylation of multiple imprinted loci (HIL) in a small proportion of patients.17
The majority of children with SRS have both birth weight and postnatal height ≤−2 SDS. In contrast to children with IUGR resulting from placental insufficiency, patients with SRS typically show no postnatal catch-up growth.3 However, there is considerable variability in the severity to which growth parameters are affected. In a recent study,17 only 78% of SRS children with confirmed molecular abnormalities studied had a birth weight ≤−2 SDS and a wide range of birth weights was observed, particularly with ICR1 hypomethylation. Children with mUPD7 have a significantly higher birth length compared with ICR1 hypomethylation patients.15 However, mUPD7 patients are more likely to show postnatal reduction in height SDS.15 On the other hand, postnatal catch-up growth can sometimes occur, particularly in children with ICR1 hypomethylation.17
Classical features of SRS include a triangular-shaped face, frontal bossing, downturned corners of the mouth and micrognathia4 (figure 2). The higher frequency of relative macrocephaly and frontal bossing tends to make the face of patients with ICR1 hypomethylation more distinctive.14 17 A triangular facial shape is more commonly observed with mUPD7.14 17 However, some patients have few, if any, of these signs and it is well recognised that the facial features of SRS tend to become less obvious with age. Photographs from early childhood may help with diagnosis in older patients.
Asymmetry can affect the trunk, face and/or limbs and is seen in around one third of all patients.4 In the cohort studied by Price et al, the maximum leg length difference was 2.5 cm.4 In older patients, limb length difference was typically 2 cm and little or no progression was seen during childhood. Asymmetry is significantly more common in association with ICR1 hypomethylation14 16 17 and may reflect mosaicism for hypomethylation at the tissue level.
Feeding difficulties in early childhood are a major feature of SRS and parents of affected children frequently comment on their lack of interest in sucking and absence of hunger from birth. Gastro-oesophageal reflux is reported in 34%, oesophagitis in 25% and food aversion in 32%.21 In a cohort of 64 SRS patients with known molecular abnormalities recently described by Wakeling et al,17 feeding difficulties were the most commonly recorded feature (seen in 86% overall), with many needing prolonged nasogastric or gastrostomy feeding. These feeding problems tend to resolve gradually over the first few years of life. By school age most SRS children feed normally, although their appetite may remain small.
Excessive sweating and hypoglycaemia
Excessive sweating is reported in around two thirds of patients, mainly in infancy.17 In some cases there is also documented evidence of hypoglycaemia. No significant difference has been found in the frequency of excessive sweating or hypoglycaemia in association with mUPD7 or ICR1 hypomethylation.17
Psychomotor retardation has been reported in just over one third of patients with SRS.4 17 22 Global delay is mostly mild and may not become apparent until mid to late childhood.4 Patients with mUPD7 are more likely to be reported as having significantly delayed development and to have received a statement of education.17
Specific problems with speech delay have been described in association with mUPD7 and referral for speech therapy is more likely to be needed by this group of children.10 17 This has been linked to the absence of paternal FOXP2 expression, as seen in other patients with developmental verbal dyspraxia.23 Early motor delay is also relatively common in children with SRS and is probably related to a combined effect of low muscle bulk and relatively large head size in infancy.17
Congenital anomalies are described in a significant minority of SRS patients. These include cleft palate, congenital heart disease, genital anomalies and limb defects.7 Major congenital anomalies appear to be much more suggestive of, though not exclusive to, ICR1 hypomethylation. Around 20% of SRS patients have generalised camptodactyly with arthrogryposis of the terminal interphalangeal joints.4 17 Fifth finger clinodactyly is significantly more common than in the general population and, along with other joint contractures, is particularly frequent in association with ICR1 hypomethylation.13 17
Myoclonus-dystonia in mUPD7
Evidence is emerging that patients with mUPD7 are at increased risk of developing myoclonus-dystonia.17 24 The disorder is associated with paternally derived mutations in the imprinted gene ε-sarcoglycan (SGCE) on chromosome 7q21. Patients with this condition typically develop mild dystonia (such as cervical dystonia or writer's cramp) and/or myoclonic jerks, especially of the arms and axial muscles, with onset of symptoms before adulthood. In one recent study, 15% of mUPD7 patients reported suggestive symptoms. These may not have previously been observed with mUPD7 as they are relatively mild and tend to develop in later childhood.
Assisted reproductive technology
Several recent studies have reported an increased frequency of assisted reproductive technology (ART) conceptions in children with other imprinting disorders including BWS or Angelman syndrome (AS).25 These findings are consistent with reports of imprinting defects in animal studies following in vivo embryo culture. Some evidence also exists for an increased frequency of ART in association with SRS, although numbers affected are small and further studies are required to confirm this.17 It is not yet clear whether this apparent increase results from an effect of embryo culture and/or a common mechanism for infertility and imprinting disorders.
In around 30–50% patients with a clinical diagnosis of SRS, the underlying molecular defect remains unknown. Recent studies have shown that a small proportion of these patients may have cryptic chromosome rearrangements detectable with microarray analysis.26 However, those patients in whom these have been found are generally less typical, often labelled as having mild SRS. This observation reinforces the importance of careful clinical assessment to try to reduce the heterogeneity in those patients labelled as having idiopathic SRS. Work is ongoing to try to determine the underlying molecular mechanism in these patients, with particular focus on the IGF/growth hormone axis and the possibility of other epigenetic mechanisms.
Several clinical scoring systems for SRS have been proposed,4 12 16 although none are well established. In 1999, Price et al described five key features (birth weight ≤−2 SD from the mean, poor postnatal growth ≤−2 SD from the mean, preservation of occipital frontal circumference, classic facial features and asymmetry) in SRS patients. Patients with classical SRS generally had at least four of these criteria.4 Following identification of ICR1 hypomethylation as an underlying cause for the condition, Bartholdi et al published criteria which they used to score 168 patients with suspected SRS.16 All patients with ICR1 hypomethylation, and seven out of 10 patients with mUPD7, fulfilled their clinical criteria for SRS.
Although the concept of a clinical scoring system is attractive, in practice a number of problems arise. Scoring for normal cognitive development is difficult in very young children, whereas facial features may be less obvious in older children. Only the scoring system proposed by Netchine et al12 includes feeding difficulties, which are a major feature of this condition. Finally, and perhaps most importantly, although application of strict scoring systems results in a high detection rate for patients with 11p15 abnormalities, there is a risk that patients with either mUPD7 or a milder phenotype associated with ICR1 hypomethylation will be missed. In the recent study by Wakeling et al,17 only 61% of patients with ICR1 hypomethylation and 20% of mUPD7 cases would have been diagnosed as haven classical SRS according to the clinical criteria of Price et al.4 No clinical feature was present in all cases and even low birth weight was present in only 78% overall. Clinicians should therefore have a low threshold for investigation of patients with features suggestive, but not typical, of SRS.
Molecular testing for 11p15 hypomethylation and mUPD7 is now widely available. If these results are normal, molecular karyotyping (microarray analysis) may also be helpful.
In children who test negative, it is important to keep an open mind about the clinical diagnosis, particularly if additional, less typical features are present. A recent review by Hall27 gives an excellent overview of the differential diagnosis in children with severe IUGR and postnatal short stature.
Prognosis and management
Management of children with SRS may require input from many different specialist teams including paediatric endocrinology for monitoring of growth and consideration of growth hormone treatment, dieticians for advice regarding food intake, orthopaedic surgery regarding limb asymmetry and clinical genetics for advice on molecular testing and recurrence risks.
The numbers of SRS patients reported in the literature who have reached their final adult height is still small and increasingly children are being treated with growth hormone, making it difficult to study the natural history of the condition. Growth studies in 386 cases found that bone age is usually delayed in early and mid-childhood, with catch-up at around 10 years, indicating an early/prompt puberty in some children. This in turn leads to relative reduction in final adult height SDS. Mean adult height in males was 151.2 cm (SD −3.7; n=11) and 139.9 cm in females (SD −4.2; n=29). Final height in patients without growth hormone treatment has been recorded in two females with confirmed ICR1 hypomethylation (142.5 cm and 146.5 cm) and one male with mUPD7 (154.0 cm).17
Growth hormone therapy
Recombinant growth hormone is now widely accepted as treatment for children born small for gestational age (SGA), including those with SRS.28 In the UK, NICE guidelines license treatment with growth hormone for children ≥4 years old with postnatal short stature (current height SDS −2.5 and parental adjusted height SDS −1) in children born with a birth weight and/or length <−2 SD who have failed to show catch-up growth. Many children with SRS will meet these criteria for treatment. Where a confident diagnosis of SRS has been made, some endocrinologists would consider starting treatment earlier, particularly in children with troublesome hypoglycaemia.
There are, however, no randomised controlled trials and only a limited number of reports on the effectiveness of growth hormone therapy specifically in SRS. Furthermore, the majority of these reports are based on small numbers of patients in whom molecular testing has not been systematically carried out.
There is some evidence that, compared with other SGA children, those with SRS respond less well to growth hormone. However, the diagnosis of SRS appears to have less effect on growth hormone response than other factors such as height SDS at the start of treatment, response to treatment in the first year, maternal height SDS and birth length SDS.29 A recent study of final height in 26 children with SRS treated with long-term growth hormone (median 9.8 years) showed a significant improvement in growth with a final height of −1.3 SDS.30 A greater increment in final height was observed in those patients with lower heights at start of treatment.
In contrast to children with mUPD7 or idiopathic SRS, those with ICR1 hypomethylation have inappropriately high serum IGF-I and IGFBP3 levels suggesting IGF-I insensitivity.15 In keeping with this finding, a trend towards more height gain on growth hormone therapy has been noted in children with mUPD7 than in those with ICR1 hypomethylation.15 Nevertheless, there are documented cases of adults with ICR1 hypomethylation who had long-term growth hormone treatment and achieved final heights well within the normal range.11 17 The possible differential effect of growth hormone in these two subgroups and idiopathic SRS is an important clinical question which deserves more extensive and prospective investigation.
Other possible benefits of growth hormone treatment, including changes in lipid profile, increase in bone mineral density, behavioural changes and improvement in self-perception, need to be carefully weighed against potential side effects, including the need for regular injections. Children with SRS being treated with growth hormone may also develop insulin resistance, but there is no evidence to date to suggest this leads to long term adverse effects on glucose metabolism.
Long term follow-up
The small numbers of adults reported with SRS have few, if any, medical problems. As a result the majority of individuals with SRS are not routinely followed up and little information is available about the natural history of the condition in adulthood. Many studies have shown that children who are SGA at birth are prone in later life to develop problems including diabetes, hypertension, hypercholesterolaemia, heart disease, hypercoagulability and osteoporosis.27 Longer term endocrine studies, looking for evidence of metabolic abnormalities in older SRS patients, will therefore be of interest.
SRS usually occurs sporadically, although a few reports of familial SRS exist in the literature, associated with various modes of inheritance.6 A small number of families have been described where more than one family member has been shown to have ICR1 hypomethylation.16 17 The underlying molecular mechanism for recurrence in these families remains uncertain. However, the vast majority of cases of SRS are non-familial and the recurrence risk remains low. Similarly, the offspring risk for affected individuals is predicted to be low.
Some clinical differences can be seen between SRS patients with mUPD7 and those with ICR1 hypomethylation. However, there is a wide range of severity and it is difficult to distinguish these two subgroups on clinical grounds alone. Clinicians should therefore have a low threshold for investigation of patients with features suggestive of SRS. There are a significant number of patients with clinical features of SRS in whom the underlying molecular mechanism remains to be determined. Our understanding of the natural history of SRS and role of treatments such as growth hormone will continue to develop as we unravel the underlying mechanisms for growth failure in this genetically heterogeneous condition.
The author would like to thank the mother of the patient shown in figure 2 for her kind permission to use these photographs.
Competing interests None.
Patient consent Obtained.
Provenance and peer review Commissioned; externally peer reviewed.