Prenatal features of Noonan syndrome: prevalence and prognostic value
Noonan syndrome (NS) is a common autosomal dominant developmental disorder, mainly characterized by congenital heart defects, short stature, and a variable degree of developmental delay. We have reviewed the prenatal findings in NS and we have correlated them with genotype and postnatal phenotype.
The cohort consisted of 47 patients with molecular diagnosis of NS. Prenatal and postnatal phenotypes were assessed by analysis of medical records, and clinical follow-up. Postnatal clinical phenotype, congenital heart disease, neuropsychomotor development, and growth pattern were arbitrarily scored in terms of severity.
Mean age at diagnosis of NS was 7 years (ranging from birth to 38 years). Abnormal maternal serum triple screen was present in 36% of cases, nuchal translucency > 2.5 mm in 41%, polyhydramnios in 38% and fetal anomalies at prenatal ultrasonography in 21%. No statistical association was observed between prenatal findings and NS genotype or scores of postnatal clinical phenotype, congenital heart disease, neuropsychomotor development, or short stature. Presence of morphologic fetal anomalies at ultrasonography was associated with developmental delay/intellectual disabilities (p < 0.001) and juvenile myelomonocytic leukaemia (p = 0.006).
Abnormal prenatal findings are frequent in NS pregnancies, though they are not specific and most are not useful for the prediction of the postnatal phenotype. Copyright © 2011 John Wiley & Sons, Ltd.
Prenatal detection of syndromic and polymalformative patterns is a continuously evolving field and a challenge for gynaecologists and clinical geneticists. Triple test, measurement of fetal nuchal translucency (NT) and morphologic ultrasound (US) are commonly used for the assessment of fetal health and screening for genetic and/or developmental defects. Noonan syndrome (NS, OMIM #163 950) is a common (1:1000–2500 live births) clinically variable disease, characterized by distinctive facial features, short stature and skeletal anomalies, congenital heart defects (CHD), cryptorchidism, lymphatic dysplasia, and a variable degree of developmental delay/intellectual disabilities (DD/ID). NS is transmitted as an autosomal dominant trait, and it is genetically heterogeneous, which partially explains its striking clinical variability. Up to date, heterozygous mutations in eight genes (i.e. PTPN11, SOS1, KRAS, RAF1, BRAF, NRAS, SHOC2 and CBL) have been shown to cause NS and closely related phenotypes (Tartaglia et al., 2001; Carta et al., 2006; Schubbert et al., 2006; Pandit et al., 2007; Razzaque et al., 2007; Roberts et al., 2007; Tartaglia et al., 2007; Sarkozy et al., 2009; Cordeddu et al., 2009; Cirstea et al., 2010; Martinelli et al., 2010; Niemeyer et al., 2010). These genes encode for proteins that participate in the RAS/mitogen activated protein kinase (MAPK) signal transduction cascade, which has been documented to play a keyrole in the control of developmental processes and growth. Remarkably, deregulation of this signalling pathway has been discovered to cause other disorders clinically related to NS, including Costello syndrome, cardiofaciocutaneous syndrome (CFCS), LEOPARD syndrome, neurofibromatosis type 1 and Legius syndrome, which together with NS are now considered to belong to a single family of disorders, the neuro-CFCSs or RAS-opathies (Tartaglia and Gelb, 2010).
Before 2001, NS diagnosis was exclusively clinical. While a comprehensive scoring system has been developed to aid in diagnosis (van der Burgt et al., 1994), clinical assessment does not usually include prenatal features, although they are frequently observed in this syndrome. Several authors have suggested that, in absence of karyotype abnormalities, this syndrome should be considered in the differential diagnosis of foetuses presenting with increased NT, especially when cardiac defects, polyhydramnios, and/or multiple effusions are concomitantly observed (Benacerraf et al., 1989; Donnenfeld et al., 1991; Nisbet et al., 1999; Achiron et al., 2000; Hiippala et al., 2001; Schluter et al., 2005; Houweling et al., 2010). The prevalence of prenatal anomalies in NS and their correlation with genotype and postnatal phenotype, however, has not been systematically investigated so far. The aim of the present study is to provide retrospective data on this matter.
SUBJECTS AND METHODS
The cohort consisted of 47 patients, including 26 (55.3%) males and 21 (46%) females, with clinical features within the NS phenotypic spectrum. Mean age at diagnosis was 7.0 years (ranging from birth to 38 years).
In all the patients enrolled in this retrospective study, diagnosis was confirmed molecularly by analyzing the entire PTPN11, SOS1, KRAS, BRAF, RAF1 and SHOC2 coding sequences, as previously reported (Tartaglia et al., 2002; Carta et al., 2006; Pandit et al., 2007; Tartaglia et al., 2007; Cordeddu et al., 2009; Sarkozy et al., 2009). The majority of the subjects were heterozygous for mutations affecting the PTPN11 (N = 30, 68% of cases) or SOS1 (N = 8, 17%) gene, while BRAF, RAF1 and SHOC2 were identified to be mutated in two subjects (4% of cases, each gene), and a single case carried a heterozygous mutation in KRAS. Most patients included in the study were sporadic cases inheriting a de novo germline mutation. Members from three families were also considered. In the first family, two sisters carried the common NS-causing c.188 A > G (Tyr63Cys) change (both parents refused the test). In the second family, two brothers inherited from the affected mother the c.854 T > C (Phe285Ile) change, that has been previously documented in NS. Finally, two sisters were found to inherit a c.1162 T > A (Leu261His) substitution from the father and the paternal grandfather, both exhibiting a mild phenotype which was only suggestive for NS.
This sequence variant was not observed among 96 population-matched unaffected individuals and over 200 patients with a clinical diagnosis of NS with normal PTPN11 sequence, suggesting a causal link with the disorder. Karyotype analysis was normal in all patients, with the exception of a familial balanced translocation [45, XX der(13,14) (q10;q10)] observed in a female proband, in her father and in her healthy brother. Informed consent was obtained from all patients included in the study.
Prenatal and perinatal phenotype assessment
Prenatal and perinatal data were collected by careful analysis of medical records and anamnestic investigation, including: results of prenatal genetic screening tests analyses (triple test and NT), occurrence of polyhydramnios, morphologic fetal US anomalies (hydrothorax, multiple effusions, malformations). In Italy, the first fetal US scan is usually performed at 11–13.6 weeks pregnancy as recommended by the Official 2010 Guidelines of the Italian Society of Gynecology and Obstetrics. An NT of 2.5 mm or greater was considered abnormal. A second and third US are usually performed at 20–22 and 30–32 weeks of gestation.
Postnatal phenotype assessment
Postnatal phenotype was assessed as follows: (1) occurrence, age of clinical onset, and severity of CHD; (2) growth pattern; (3) neuropsychomotor development; (4) electroencephalography (EEG) anomalies, and/or epilepsy and (5) occurrence of haematological anomalies.
To grade and standardize the clinical severity of the phenotype, a scoring system assigning one and three points respectively to the minor and major van der Burgt's criteria (van der Burgt, 1994) was developed, allowing to classify the phenotype as mild (score = 1–6), moderate (score = 7–12), or severe (score = 13–18). A second standardized scoring system was used to classify CHD, where score 0 refers to normal cardiac development, whereas score 4 refers to severe or lethal CHD. A similar method was applied to define neuropsychomotor development, ranging from normal achievement of developmental milestones and learning abilities (score 0) to significant speech delay associated with DD/ID (score 3). Short stature was graded as severe with height below the third percentile (score 2), moderate between the third and tenth percentile (score 1) or absent above the tenth percentile (score 0; Table 1).
Table 1. Scoring system
|Clinical phenotype (based on van der Burgt's criteria)||1–6 = mild|
|Three points for each major featurea , one point for each minor featureb||7–12 = moderate|
| ||13–18 = severe|
|CHD||0 = absent|
| ||1 = spontaneous resolution or not specific treatment required|
| ||2 = pharmacological treatment required|
| ||3 = surgical treatment required|
| ||4 = not resolutive surgery or heart transplantation or death|
|Neuropsychomotor development||0 = normal|
| ||1 = delayed achievement of developmental milestones|
| ||2 = learning disabilities|
| ||3 = learning disabilities and speech impairment|
|Growth pattern||0 = normal|
| ||1 = stature < 10 percentile|
| ||2 = stature < 3 percentile|
Associations between prenatal features, genotype, and postnatal phenotype were explored among all the variables considered. Fisher's exact test or Chi square (χ2) test was employed for nominal variables. A p-value less than 0.05 was considered statistically significant.
Our study group consisted of 47 patients with prenatal information, postnatal phenotype and molecular analysis available. Prenatal US had been performed in the entire patient group, whereas triple screen and measurement of NT were also determined in 22 subjects (46.8%). Among the prenatal findings, abnormal results at triple screen were reported in 8/22 cases (36%), increased NT in 9/22 (41%), polyhydramnios in 18/47 (38%), and abnormal sonographic fetal findings in 10/47 (21%). Any abnormal prenatal finding was present in 20/47 cases (43%). Table 2 summarizes the associations between prenatal features in the cohort and the postnatal scores of disease severity.
Table 2. Association between prenatal findings and postnatal scores
|Clinical phenotype||p = 0.979||p = 0.899||p = 0.158||p = 0.362|
|CHD||p = 0.924||p = 0.771||p = 0.344||p = 0.169|
| Score 0||1||1||1||1||2||4||0||6|
| Score 1||5||9||5||9||11||12||6||17|
| Score 2||0||1||0||1||0||2||0||2|
| Score 3||1||1||2||1||3||10||2||11|
| Score 4||1||2||1||1||2||1||2||1|
|Neuropsychomotor development||p = 0.466||p = 0.293||p = 0.003||p = 0.009|
| Score 0||1||5||1||5||3||19||0||22|
| Score 1||6||6||7||5||11||6||7||10|
| Score 2||1||2||1||2||2||4||2||4|
| Score 3||0||1||0||1||2||0||1||1|
|Growth pattern||p = 0.329||p = 0.530||p = 0.544||p = 0.584|
| Score 0||2||1||2||1||4||5||1||8|
| Score 1||1||5||2||5||3||9||2||10|
| Score 2||5||8||5||7||11||15||7||19|
Table 3 presents full molecular and clinical characterizations of the 10 patients presenting morphologic fetal US anomalies. Hydrothorax was the most frequent morphologic anomaly, occurring in five of them. Genotype–phenotype correlations did not give evidence of a different distribution of prenatal anomalies among the NS specific-gene subgroups.
Table 3. Genotype, prenatal and postnatal phenotype in patients with morphologic fetal US anomalies
|1||PTPN11||▪ Triple test not performed||▪ Type III CHD (ASD + moderate||11 (moderate)|
| ||Asp61Asn||▪ Increased NT|| mitral incompetence)||3 major + 2 minor|
| || ||▪ Polyhydramnios||▪ Stature < 10 centile|| |
| || ||▪ Hydrothorax||▪ DD/ID (class 1)|| |
| || ||▪ CHD (ASD)||▪ JMML/MDS|| |
|2||PTPN11||▪ Abnormal triple test||▪ Type I CHD (elctrocardiographic||9 (moderate)|
| ||Asp61Asn||▪ Increased NT|| anomalies)||3 major|
| || ||▪ Polyhydramnios||▪ Stature < 3 centile|| |
| || ||▪ Renal anomalies (right||▪ DD/ID (class 1)|| |
| || || kidney pyelectasia)||▪ JMML|| |
|3||PTPN11||▪ Abnormal triple test||▪ Type I CHD (ASD)||5 (mild)|
| ||Gln79Arg||▪ Increased NT||▪ DD/ID (class 1)||2 major + 2 minor|
| || ||▪ Polyhydramnios|| || |
| || ||▪ Pleural effusion|| || |
| || ||▪ Central nervous system|| || |
| || || anomalies (dilation|| || |
| || || fourth ventricle, absence|| || |
| || || cauda cerebelli)|| || |
|4||PTPN11||▪ Triple test not performed||▪ Type I CHD (PS + ASD)||13 (severe)|
| ||Glu139Asp||▪ NT not performed||▪ Stature < 3 centile||4 major + 1 minor|
| || ||▪ Polyhydramnios||▪ DD/ID (class 2)|| |
| || ||▪ Hydrothorax and pleural||▪ Epilepsy|| |
| || || effusions||▪ JMML|| |
|5||PTPN11||▪ Normal triple test||▪ Type IV CHD (HCM + PS)||12 (moderate)|
| ||Phe285ser||▪ Normal NT||▪ Stature < 3 centile||4 major|
| || ||▪ Polyhydramnios||▪ DD/ID (class 1)|| |
| || ||▪ Hydrothorax||▪ JMML|| |
|6||BRAF||▪ Abnormal triple test||▪ Type I CHD (patent foramen ovale)||11 (moderate)|
| ||Gln257Arg||▪ Increased NT||▪ Stature < 3 centile||3 major + 2 minor|
| || ||▪ Polyhydramnios||▪ Ectodermal anomalies|| |
| || ||▪ CHD (left superior vena cava||▪ DD/ID (class 3)|| |
| || || in coronary sinus)|| || |
|7||BRAF||▪ Triple test not performed||▪ Type I CHD (ASD)||9 (moderate)|
| ||Leu597Val||▪ NT not performed||▪ Stature < 3° centile||2 major + 3 minor|
| || ||▪ Polyhydramnios||▪ DD/ID (class 3)|| |
| || ||▪ Renal anomalies (left||▪ CNS anomalies (Chiari type I)|| |
| || || kidney pyelectasia)||▪ Epilepsy|| |
|8||RAF1||▪ Normal triple test||▪ Type I CHD (PS)||12 (moderate)|
| ||Pro261Ser||▪ Normal NT||▪ Stature below < 3° centile||4 major|
| || ||▪ Polyhydramnios||▪ DD/ID (class 1)|| |
| || ||▪ Intra-uterine growth restriction|| || |
|9||SOS1||▪ Abnormal triple test||▪ Type III CHD (PS + VSD)||14 (severe)|
| ||Met269Thr||▪ Increased NT||▪ Stature < 10 centile||4 major + 2 minor|
| || ||▪ Polyhydramnios||▪ DD/ID (class 2)|| |
| || ||▪ Hydrothorax and CHD (VSD)||▪ Epilepsy|| |
|10||SHOC2||▪ Abnormal triple test||▪ Type IV CHD (HCM + VSD +||15 (severe)|
| ||Ser2Gly||▪ Increased NT|| dysplastic mitral valve + PS)||5 major|
| || ||▪ CHD (left superior vena||▪ Stature < 3° centile|| |
| || || cava in coronary sinus||▪ DD/ID (class 3)|| |
| || || and moderate tricuspid||▪ Epilepsy|| |
| || || incompetence)||▪ MDS|| |
Overall, 50 CHD were documented in 41 patients (87.2% of the study cohort): 30 pulmonic stenosis, 7 hypertrophic myocardiopathy, and 13 septal defects. However, only in four cases the CHD has been detected prenatally (9.7%): septal defects were observed in two patients, left superior vena cava in coronary sinus in the other two. On the basis of the scoring system referred to CHD, 23 subjects (56%) exhibited spontaneous resolution of the anomaly or did not required specific treatment (score 1), two individuals (4.8%) required pharmacological treatment (score 2), while in 13 cases (31.7%) a surgical treatment was necessary (score 3), and in three subjects (7.3%) surgery was not resolutive or the severity of the disease required heart transplantation or was the cause of death (score 4). No statistically significant association was observed between CHD and prenatal findings.
Normal neuropsychomotor development—score 0—was observed in 23 patients (49%). Delayed achievement of early neuromotor milestones with subsequent normal development (score 1) was observed in 15 subjects (31.9% of the entire cohort), DD/ID (score 2) was observed in six individuals (12.7%), while in three patients (6.3%) DD/ID was associated with a severe speech delay (score 3). A significant statistical correlation was found between morphologic fetal US anomalies and DD/ID (scores 1–3) (p < 0.001), even though the statistical power of this association decreased when only significant neurodevelopmental anomalies (scores 1–2) were taken into consideration. No associations were observed between pathologic triple test and/or increased NT and neuropsychomotor score.
Severe reduced growth occurred in 26 patients (55.3%), while moderate short stature was observed in 11 subjects (23.4%). No association was present between growth pattern and prenatal findings.
Five patients (11%) presented haematological anomalies, with juvenile myelomonocytic leukaemia (JMML) documented in four of them, and a myelodysplastic disorder in one, spontaneously resolved in all of them. Morphologic fetal US anomalies were more frequent among NS children with haematologic anomalies than without them (5/5 vs 5/42, p < 0.001).
Prenatal features of NS and its prenatal diagnosis have been reported in several studies, but a clear correlation between specific prenatal features and the postnatal phenotype has not been investigated thoroughly. Several authors suggested that NS might be suspected when cystic hygroma, ascites or pleural effusion, hydrops fetalis, polyhydramnios, increased NT or CHD are observed (Benacerraf et al., 1989; Donnenfeld et al., 1991; Nisbet et al., 1999; Achiron et al., 2000; Hiippala et al., 2001; Schluter et al., 2005; Houweling et al., 2010). Sharland et al. (1992) reported polyhydramnios in the prenatal history of 33% of NS patients. The aetiology of this feature is still unclear, but it could be related to fetal swallowing anomalies. In a recent report (Lee et al., 2009), it has been estimated that PTPN11 mutations are identifiable in 2 and 16% of fetuses presenting increased NT and cystic hygroma, respectively, confirming the poor specificity of these findings.
As far as we are aware, this is the first report detailing the prenatal features of NS retrospectively by investigating a clinically well-characterized and relatively large cohort of patients with diagnosis confirmed molecularly.
This report supports the view that a significant subset of subjects with NS present with a wide variable range of prenatal anomalies, the most frequent being polyhydramnios (38.3%), and pathologic triple test and increased NT (36.4%). Of note, these prenatal findings do not correlate with any investigated aspect of the postnatal phenotype, and the overall prognosis of the syndrome. Morphologic fetal US anomalies were detected in more than one fifth of subjects, with the most common being hydrothorax, observed in 10.6% of cases. Among these subjects, only 3 of the 10 patients with fetal morphologic US anomalies were characterized by a severe postnatal phenotype. Specifically, five of these subjects presented class 1 DD/ID with good prognosis, while five of them presented a more severe DD/ID (class 2, N = 2; class 3, N = 3). Interestingly, all the five patients presenting myelodysplastic syndrome (MDS)/JMML were characterized by a pathologic prenatal history characterized by morphologic fetal US anomalies, with three of them presenting hydrothorax. The correlation between fetal multiple effusions and MDS/JMML is intriguing and deserves further analysis. Of note, most of the prenatal parameters investigated were unable to predict the postnatal outcome disclosing a very poor correlation with prognosis. Anyway, prenatal US anomalies, mostly hydrothorax, were observed to correlate to an increased likelihood of MDS/JMML and DD/ID.
The observed timing of diagnosis of CHD in the present cohort of patients, a key feature of NS, deserves specific comment. Specifically, only a minor fraction (8%) of CHD observed in this cohort was diagnosed prenatally, whereas approximately half of them were clinically evident at birth.
This study carries the limitation of the retrospective design, being unblinded and prone to recall bias. It should be also considered that to the accuracy of the study and to better evaluate eventual genotype–phenotype correlations, we only included patients with a positive molecular testing. As a consequence, the majority of subjects were PTPN11 mutation-positive patients. While no evidence of a significantly different distribution of prenatal anomalies among patients with mutations in different disease genes was apparent, specific correlations between prenatal and postnatal features in patients with mutations in genes weakly represented in the present cohort (i.e. SHOC2, KRAS and BRAF) cannot be ruled out.
In conclusion, this study represents a retrospective investigation about prenatal phenotype in NS, and its correlation with the severity of the prognosis. Even though non-specific, prenatal anomalies are very frequent in NS pregnancies. Our results also evidence that diagnosis of cardiac anomalies is rarely made prenatally. Finally, although most of the prenatal features we described resulted not useful for the prediction of the phenotype evolution, some correlations between prenatal anomalies and postnatal phenotypic features are quite interesting and deserve further studies on larger cohorts to better define their meaning and to better understand their clinical implications.
We acknowledge Dr Faustina Lalatta for suggestions and critical reading of the manuscript. We thank Regione Piemonte, Progetto Ricerca Sanitaria Finalizzata for financial support.