Introduction

Disturbances of the chromosomal region 11p15.5 are associated with Beckwith–Wiedemann (BWS, MIM130650) and Silver–Russell syndromes (SRS, MIM180860) (Table 1). Both congenital disorders belong to the group of imprinting disorders (IDs), that is, entities that are characterized by molecular alterations at the imprinted loci. These specifically regulated genes are expressed either only from the maternal or from the paternal gene copy but not biparentally (for a review, see Reik and Walter1). Many of the currently known 100 imprinted genes are found in clusters, that is, the imprinted loci often comprise multiple genes under coordinated control. This control is mediated by differentially methylated regions (DMRs), stretches of CpG islands in which the cytosine residues are methylated in a parent-of-origin manner. So far, nine IDs have been defined, and they are associated with changes in seven imprinting clusters (for a review, see Mackay et al2 and Eggermann et al3).

Table 1 Molecular findings and major prenatal and postnatal features in the 11p15-asscociated IDs
Full size table

Among the imprinted loci, the chromosomal region 11p15.5 has a key role4, 5 in the etiology of BWS and SRS. It harbors two separate imprinting control regions (ICRs): the imprinting control region 1 (ICR1) is methylated on the paternal allele and regulates the maternally expressed H19 and paternally expressed IGF2 genes via the H19 differentially methylated region (H19-DMR), whereas the maternally methylated KCNQ1OT1-DMR in the ICR2 regulates the maternally expressed CDKN1C and paternally expressed KCNQ1OT1 genes.

In BWS, both epimutations and genomic alterations affecting 11p15.5 are present in >70% of patients, whereas up to 60% of SRS patients carry 11p15.5 disturbances (Table 1). Additionally, 10% of the latter patients show a maternal uniparental disomy of chromosome 7 (upd(7)mat). Furthermore, in patients exhibiting features of the two disorders aberrations not involving chromosomes 7 and 11p15.56, 7, 8, 9 are reported.

Because of the methodological improvement and the growing knowledge on 11p15-associated IDs, there is an increasing demand not only for postnatal confirmation of a clinical diagnosis but also for prenatal testing. This request might also be boosted by an altered reproductive behavior at least in developed countries: both advanced maternal age at childbirth as well as assisted reproduction technologies (ART) have been suggested as predisposing factors for IDs. Advanced maternal age might promote the development of chromosomal aberrations, including uniparental disomy,10, 11 whereas the ART procedure itself might influence the establishment and/or maintenance of imprinting marks.12

However, the complex molecular findings and alterations in BWS and SRS patients are a challenge for laboratories and clinical geneticists offering these tests. They have to consider a wide range of detectable disturbances, occurrence of mosaicism, limitations of the applicable tests, consequences for the patients, the families and – in the case of prenatal testing – the informative value of the test results. Based on their experiences and work in the field of 11p15.5-associated disorders, the authors of this paper are aware of these challenges and summarize the current knowledge on prenatal testing in BWS and SRS.

Molecular alterations in IDs and mosaicism

Similar to other IDs, distinct molecular sub-groups can be detected in BWS and SRS: (a) (micro)chromosomal deletions, duplications and rearrangements; (b) mutations in imprinted genes or their imprinting control regions (ICR); (c) uniparental disomy (UPD); and (d) epimutations (ie, aberrant methylation without alteration of the genomic DNA sequence).

Chromosomal or intragenic mutations affecting imprinted loci (a, b) can occur de novo, but in many cases familial inheritance has been reported. UPD (c) and epimutations (d) do not affect the genomic DNA sequence itself, but the regulation of gene expression by parent-of-origin imprinting. All subgroups affect the fine-tuned expression of genes controlled by ICRs.

When UPDs and epimutations arise postzygotically during early embryonic cell divisions, they are most likely associated with a mosaic distribution (Table 1). However, mosaicism has also been reported for genomic alterations.13 Mosaicism can disturb genotype–phenotype correlations and may explain, in part, the broad phenotypic range associated with these disorders (Table 1). From the diagnostic point of view, it can significantly hamper the detection of the molecular alteration (eg, Alders et al14). Therefore, it is in principle impossible to exclude the clinical diagnosis in case of a negative test result, because other untested tissues might carry the alteration.

Genomic mutations and epimutations in BWS and SRS

Both SRS and BWS are mainly associated with molecular alterations affecting the two ICRs in 11p15.5 (Table 1). The molecular defects in 11p15.5 in the two disorders comprise the four different types of mutations and epimutations described before (for a review, see Shuman et al15 and Saal16). In patients with SRS features, a considerable number of cases carry a upd(7)mat or (submicroscopic) changes of different chromosomes.17, 18 Furthermore, the number of SRS patients with reported molecular defects in 14q32 corresponding to Temple syndrome (TS14) is increasing.6, 7, 19 For patients with BWS features, aberrations of other chromosomal regions and genes have rarely been reported.8, 9, 20

Up to 60% of SRS patients show a DNA hypomethylation of the ICR1 in 11p15; in addition, in single cases complete or partial maternal duplications of 11p15 have been identified (for a review, Begemann et al21). Other disturbances of chromosome 11 (upd(11)mat; CDKN1C, IGF2 point mutations) are rare in SRS, and the gene variants only occur in familial cases.22, 23, 24, 25

In BWS, the opposite epigenetic or genetic changes in 11p15 have been observed in nearly 70% of patients with a preponderance of DNA hypomethylation at the ICR2 accounting for up to 50% of all BWS cases. Upd(11p15)pat is the second most common alteration (~20%), while DNA hypermethylation at ICR1 is rare (5–10%). In up to 5% of sporadic and 50% of familial BWS patients, CDKN1C point mutations can be identified. Additionally, point mutations and small deletions at protein-binding sites within the ICR126 as well as microdeletions within the ICR127 or of the complete ICR228, 29 are observed. In these families, BWS follows autosomal-dominant inheritance with incomplete penetrance dependent on the sex of the parent contributing the mutation. For both IDs, a growing number of cases and families with small rearrangements and imbalances restricted to single DMR in either the ICR1 or ICR2 have been reported.21

Despite the initial correlation between 11p15 in BWS and 11p15 and chromosome 7 in SRS, a growing number of phenotypically related patients are described displaying hypomethylation of imprinted loci additional to the disease-specific ones. This multilocus imprinting disturbance (MLID) is detectable in 7% of SRS patients with ICR1 and 25% of BWS patients with ICR2 hypomethylation. Monogenic mutations have been reported that predispose to MLID,30, 31, 32 thus the prediction of a recurrence risk is difficult in patients carrying MLID.

Clinical findings in BWS and SRS

The most prominent clinical feature in both IDs is growth disturbance. Indeed, the overgrowth in BWS and the growth restriction in SRS impressively reflect the underlying opposite molecular defects (Table 1). In both syndromes, the clinical diagnosis is often difficult owing to their variable presentation during childhood and adolescence, and the molecular and phenotypic findings overlap with other IDs and syndromes.15, 16 Thus differential diagnoses should always be considered, in particular in the course of a (prenatal) diagnostic workup.

Beckwith–Wiedemann syndrome

BWS was originally referred to as EMG syndrome because of its three cardinal features: Exomphalos, Macroglossia and (neonatal) macrosomia (Gigantism). A recognizable facial gestalt of BWS is common and may include prominent eyes with intraorbital creases, facial naevus flammeus, midfacial hypoplasia, full lower lips, earlobe creases and posterior helical pits. Additional clinical features of BWS include neonatal hypoglycemia, hemihyperplasia, organomegaly (heart, liver, spleen, pancreas, kidneys, adrenals) and unilateral or bilateral renal abnormalities (eg, medullary dysplasia, nephrocalcinosis, medullary sponge kidney) (Table 1). In 5–7% of children with BWS, embryonal tumors (most commonly: Wilms tumor, hepatoblastoma, neuroblastoma, and rhabdomyosarcoma) are diagnosed. Early death may occur from complications of prematurity, hypoglycemia, cardiomyopathy, macroglossia or tumors. Growth rate slows around age 7–8 years. The tumor risk also reduces beyond this age, but in case of upd(11)pat a mosaic genome-wide paternal uniparental diploidy33 has to be excluded as these patients have a lifelong risk to develop tumors. Hemihyperplasia may affect segmental regions of the body or selected organs and tissues. Prenatally, the major signs detectable by ultrasound are placental mesenchymal dysplasia,34 omphalocele (detectable after the thirteenth week of pregnancy (wp)), polyhydramnios and fetal macrosomia (detectable after the nineteenth wp) (Table 1).35 The incidence of polyhydramnios, premature birth and fetal macrosomia in BWS has been estimated as ~50% (for gene reviews, see Shuman et al15). Other common features include a long umbilical cord and an enlarged placenta that averages almost twice the normal weight for gestational age. During pregnancy, preeclampsia or eclampsia may occur.35, 36

Silver–Russell syndrome

SRS is characterized by intrauterine and postnatal growth restriction with a typical facial gestalt (for a review, Saal16). In the majority of patients, birth weight is 2–3 s.d. below the mean, and growth restriction persists. As a result, mean adult height without GH treatment is 150 cm in males and 139 cm in females.37 Head circumference can be normal, thus the patients show a relative macrocephaly. Clinodactyly V and face/body asymmetry (hemihypoplasia) are frequent. The typical facies include a prominent forehead, a triangular face and a pointed chin. Children with SRS might exhibit a (slight) developmental delay (both motor and cognitive) and learning disabilities, however, these features predominantly occur in carriers with upd(7)mat.38 Complications of SRS include gastrointestinal reflux and failure to thrive. Children may benefit from GH treatment. Intrauterine growth retardation and relative macrocephaly usually occur in the third trimester and are often the only clinical features of SRS to be detected by fetal ultrasonography. In rare cases, hypospadia and/or cleft palate are reported associated with a very small placenta.

Recurrence risk estimation in BWS and SRS

The majority of cases with BWS and SRS have been reported to occur sporadically; this is reflected by the type of (epi)mutations in both disorders: ICR1 hypomethylation in SRS as well as ICR2 hypomethylation and upd(11)pat in BWS mainly occur as mosaicism and probably originate from postzygotic errors. In contrast, constitutional mutations (point mutations, duplications/deletions) are associated with a significantly increased recurrence risk of up to 50% depending from the affected paternal allele. However, with our increased knowledge about molecular disturbances in both syndromes, we must adjust risk estimation and counseling. Furthermore, specific mutations have been identified that cause aberrant methylation either by acting in cis (eg, small deletions in ICR139) or in trans (NLRP232) In particular, approximately 20% of BWS patients with ICR1 hypermethylation have point mutations or small deletions in OCT4/SOX2-binding sites within the H19-DMR.26

Thus, to offer genetic counseling to families with IDs, the knowledge of the nature of the mutation or epimutation subtype is essential to delineate exact risk figures, and genetic counseling by an experienced clinical geneticist is emphasized.

Molecular testing for BWS and SRS

The currently applied molecular tests in ID diagnostics are mainly restricted to single disease-specific loci, but technical, biological and clinical factors influence the diagnostic yield. An increasing number of molecular techniques are available (Table 2) either targeting only one imprinted region or several DMRs. The current lack of standardization can make comparisons between molecular results from different laboratories problematic, furthermore some of the applied methods do not allow the discrimination between the different types of (epi)mutations (Table 2). Several laboratories performing clinical testing for BWS and SRS already participate in informal sample exchange programs or have extensively validated several different testing approaches. Additionally, the European Molecular Quality Network (EMQN) has established a BWS/SRS quality assessment scheme for 11p15.5-associated IDs.

Table 2 Suitability of different (methylation-specific (MS)) methods for prenatal testing
Full size table

A major challenge is mosaicism (eg, Azzi et al40). In case of a low-level mosaicism, a false-negative result will be generated. Thus, in case of an unequivocal diagnosis and negative testing results in DNA derived from blood, another tissue (fibroblasts, buccal cells) should be tested.

Another problem of diagnostic testing for IDs is the broad spectrum of molecular alterations: an imprinted locus can be affected by different types of (epi)mutations. Additionally in some IDs even different loci might be affected. Therefore, the diagnostic workup for each ID requires a profound knowledge on the possible molecular findings and the limitations of the applied (Tables 1 and 2). For putative variants, additional laboratory tests (eg, molecular karyotyping, FISH, genotyping) might be required; for point mutations, a bioinformatics estimation of its pathogenicity has to be performed, along with family studies to determine the mode of transmission and segregation with phenotype.

It is foreseeable that the increasing implementation of next-generation sequencing (NGS)-based assays in routine diagnostics will also find the way in the diagnostic workup of IDs. NGS has the potential to detect all types of (epi)mutations at once and can target multiple loci; massive parallel sequencing will also allow a more sensitive mosaic detection. However, future non-invasive prenatal testing (NIPT) of IDs will probably be restricted to families with chromosomal rearrangements. Prenatal DNA methylome profiling or transcriptomic analysis might indeed result in the identification of biomarkers in the early diagnosis of placental dysfunction or aneuploidies and other developmental defects,41, 42 but it might not be suitable for NIPT of IDs owing to the early time of sample drawing and the multiple challenges in determining the fetal methylome/transcriptome status in maternal plasma.

Experiences from prenatal BWS and SRS testing

In the centers contributing to this review, 160 prenatal tests have been performed (Tables 3a,b,c), 122 for BWS and 38 for SRS. In this cohort, several criteria have been evaluated, including the methods applied, the tissues analyzed (chorionic villus (CVS), amniocytic fluid cells (AF) or fetal blood; cultured and native cells), the reasons for referral, the number of positively tested cases and the number of false-positive/negative results (Tables 3a,b,c).

Table 3a Data on the prenatal cohort tested for BWS or SRS: overview on the tested tissues and the applied methods in prenatal BWS and SRS testing
Full size table
Table 3b Data on the prenatal cohort tested for BWS or SRS: tissues used for prenatal testing in BWS and SRS
Full size table
Table 3c Data on the prenatal cohort tested for BWS or SRS: reason for referral for prenatal BWS or SRS testing
Full size table

Methods

Different methods are applied (Table 3a), with different sensitivities and limitations with respect to the discrimination between the different molecular subtypes (Table 2). However, as shown for the imprinted GNAS locus, the heterogeneity of molecular tests does not affect the correctness of the molecular diagnosis.43 Thus no specific assay can be recommended, but in any case a careful validation has to be carried out and advantages and disadvantages in the context of prenatal testing (duration of the protocol, amount of DNA) have to be considered. For the methods analyzed by our centers, between 50 (MS-MLPA) and 100 (HRMA, AMMS-RTqPCR) control samples have been used for validation. Furthermore, the limitations of the different tests have to be taken into account, for example, microsatellite typing can be peformed to determine UPD, but this method does, of course, not exclude epimutations.

Prenatal tissue

Both native and cultured prenatal cells might be suitable for molecular testing (Tables 3a and b). However, it should be kept in mind that cell culture might result in clonal features that do not always correlate with the true biological status of the fetus/placenta. Furthermore, in the case of early CVS (before 12 wp) the imprinting marks at the loci of interest might not be finally set.44 As CVS is an extraembryonic tissue, it might not reflect the (epi)genetic constitution of the fetus. Although the analysis of cells derived from the fetus itself offers the best chance to detect altered imprinting marks, our data show that even in this case the risk of false-positive results cannot be prevented (Table 3b). In general, mosaicism is a major challenge in molecular testing of BWS and SRS, and it can never be excluded.

Reasons for referral for prenatal testing

As the experiences from our centers show, the reasons for referring prenatal samples for testing differ remarkably between BWS and SRS (Table 3c). A positive family history is generally a rare indication for both disorders and is generally restricted to monogenic mutations (eg, CDKN1C) or chromosomal rearrangements (eg, 11p duplications). In BWS, abnormal ultrasound (ie, abdominal wall defects/omphalocele) is the main reason for invasive prenatal testing. Although IUGR is the key prenatal clinical feature of SRS, it is a rare indication for prenatal testing, presumably because it becomes obvious only in the third trimester and is generally an unspecific symptom, which should be regarded with caution as an indication for prenatal testing. In prenatal SRS testing, chromosomal aberrations (trisomy 7 in CVS, familial translocations) are the predominant indications.

Prenatal testing results and pregnancy outcome

The results of prenatal testing for BWS and SRS reflect both the clinical heterogeneity and the molecular complexity of the disorders (Tables 3a and b). In the case of SRS, IUGR is the most obvious ultrasound sign but very unspecific and can therefore merely be regarded as a reliable feature for prenatal testing. This might explain why no positive cases have yet been ascertained prenatally. Indeed, the major reasons for prenatal testing for SRS are chromosomal aberrations, that is, familial translocations involving chromosomes 7 and 11 or trisomy 7 in CVS.

The distribution of positive test results in the cases ascertained for BWS testing reflects the published frequencies of (epi)mutations (Tables 3a and b). The ICR2 hypomethylation accounts for the majority (n=15) of the 20 positive test results, 4 fetuses carried upd(11)pat and 1 ICR1 hypermethylation. In 32 fetuses, prenatal CDKN1C sequencing has been performed. In two fetuses, the mothers had been identified as mutation carriers. In the other cases, ultrasound findings were indicative for BWS. In that group, four fetuses carried pathogenic mutations, in one the mutation was maternal in origin, too.

False-positive/negative cases

In one case initially diagnosed as ICR2 hypomethylation in CVS, the analysis of a second prenatal tissue (amniotic fluid, AF) and the clinical outcome revealed a false-positive testing result. These discrepant findings can be explained by an incomplete methylation at the time of sample drawing (11+6 wp). In two cases with a negative prenatal test result, postnatal testing identified ICR2 hypomethylation and upd(11)pat, respectively, thereby confirming the prenatal suspicion of BWS. This discrepancy is in accordance with a previous report on the mosaic distributions of molecular disturbances in BWS.14 The example vividly illustrates that laboratories offering prenatal (and postnatal) tests have to be aware of and to report on the risk to miss 11p15 defects owing to somatic mosaicism and the value of prenatal testing has to be critically discussed with the families.

Challenges for prenatal testing in BWS and SRS

In general, prior to molecular prenatal testing, genetic counseling should be offered to discuss the informativity and limitations of prenatal ID testing (Figure 1) and finally the consequences of both positive and negative test results.

Figure 1
figure 1

Considerations to be addressed prior to prenatal testing of BWS and SRS.

Full size image

First of all, general issues of prenatal (molecular) have to be considered, such as the individual situation (risk and suitability of the invasive test for the current pregnancy), the risk of contamination of the fetal sample with maternal cells resulting in a false-negative result (a general hazard of invasive prenatal testing), the suitability of the invasively drawn sample for the applied test (amount and quality of the DNA) and the difficult phenotype prediction in the case of a positive testing result.

Furthermore, ID-specific considerations have to be addressed:

  • The timing of sample drawing and the methylation status of the DMR of interest at that time. It is well known that some DMRs are not finally established during the time of CVS in the 10–12th wp,44 and methylation-specific testing at that time leads to false results, explaining the false-positive case in our cohort (Tables 3a and b).

  • Mosaicism can cause a false-negative testing result as shown in our study cohort (Tables 3a,b,c).

  • So far, there is no consensus on the target methylated CpGs and DMRs in ID testing, therefore the experiences and data are scarcely comparable between different laboratories, and every test has to be validated separately in-house.

  • Because of the heterogeneity and complexity of the molecular findings in the 11p15-associated IDs, unusual molecular alterations may elude identification. Additionally, the informative content of the available assay are different and might hamper the diagnostic process (Table 2).

Whereas these aspects address the laboratory workup and are the basis for a reliable test result, the handling of both negative and positive results are probably in the purview of the genetic counseling session. However, the technical limitations should also be included in the written report.

A positive prenatal test result

In the rare case of a positive family history, the families are generally aware of the significance of a positive testing result. The situation is different for the majority of couples who are confronted with the possibility of giving birth to a child with BWS or SRS for the first time. In this situation, the molecular workup might help to confirm a clinical diagnosis and exclude other entities, in particular those associated with intellectual disabilities. In all such pregnancies, detailed ultrasound at least at 18–20 wp and again at 25–28 wp should be directed to assess growth parameters and the presence of abdominal wall defects, organomegaly, renal and cardiac abnormalities, cleft palate and macroglossia. Early prenatal confirmation of a clinical diagnosis can result in recommendation for hospital delivery, monitoring the baby for hypoglycemia, surveillance for tumors, further therapeutic options and support of the parents. Both BWS and SRS babies may have feeding difficulties in the neonatal and early childhood period.

Nevertheless, in some circumstances a positive test result may lead to a decision to terminate the pregnancy. Vice versa, a positive diagnostic test might result in a therapeutic abortion being avoided – for example, in a fetus with an anterior wall defect the molecular diagnosis of BWS implies that the risk of learning disability is small. Indeed, some of the contributing authors are aware of pregnancies with a negative molecular testing, which were then terminated because of the uncertainty of the underlying cause of the clinical finding.

A negative prenatal test result

As already discussed, the major molecular alterations both in BWS and SRS (ie, epimutations in SRS/BWS and upd(11)pat in BWS) can occur in the mosaic form14, 21 and might escape molecular detection because of the low level of cells harboring the disturbance. The authors are aware of two false-negative prenatal testing result with an epimutation detected after birth (Table 3b). Hence, a normal prenatal test result cannot absolutely exclude a diagnosis of BWS and SRS. Furthermore, as previously indicated, unknown genes and mechanisms are likely to exist in addition to the mosaic cases that escape detection using current testing approaches. These issues need to be emphasized in genetic counseling and in clinical reports.

Postnatal testing

In case of a persisting clinical suspicion of SRS and BWS after a negative prenatal testing result, postnatal testing should also be performed to confirm any findings. Indeed, the limitations of prenatal testing need to be very clear prior to testing being undertaken.

Conclusion

The growing knowledge on the molecular basis of BWS and SRS and the increasing number of positively tested patients results in an increasing demand for prenatal testing for these diseases. Before offering these tests, numerous questions have to be tabled and discussed with the couple, ranging from methodological questions to the ethical topics as listed in Figure 1. In particular, in case of a negative testing result, its informative value has to be critically discussed. However, the optimization of tests and their implementation as well as the current quality assessment schemes under development by the EMQN for BWS and SRS will contribute to a more reliable and meaningful prenatal diagnosis. Thus the decision on the continuation of a pregnancy suspicious for BWS or SRS has to take into account all these considerations, but finally it has to be reached by the parents after a self-determined decision process and according to the national laws.