Noonan Syndrome Growth Charts and Genotypes: 15-Year Longitudinal Single-Centre Study

Noonan syndrome (NS) is a relatively common genetic condition with an estimated incidence of one in 1,000–2,500 people [1, 2]. It is characterised by clinical features such as short stature, facial dysmorphism, congenital cardiac abnormalities, reduced growth, chest deformity, cognitive deficits, cryptorchidism, delayed puberty, and lymphatic dysplasia [1, 3‒7]. Congenital cardiac abnormalities are most common, occurring in up to 80% of cases, and often include pulmonary valve stenosis (PS) and hypertrophic cardiomyopathy (HCM) [1, 4, 8].

NS is primarily inherited as an autosomal dominant trait caused by a number of genetic variants, most commonly involving the RAS/MAPK pathway [1, 5, 6, 9, 10]. These variants belong to a group of diseases called RASopathies [6, 8, 9], about which little is known due to their heterogeneity in clinical presentation and genetic variation [9]. Effects on the endocrine system (thyroid function), metabolism, and puberty [2, 11, 12] are yet to be documented.

In NS, PTPN11 gene mutations account for almost 50% of cases, with SOS1 mutations accounting for 11% of cases, as well as RAF1 and RIT1 mutations (5%), KRAS, NRAS, SHOC1, CBL, LZTR1, BRAF, SOS2, and MAP2KI mutations, with approximately 20% remaining unknown [4, 7, 8, 13‒16]. The typical NS short stature phenotype seems to correlate more with PTPN11 mutations than with SOS1 mutations [6, 17], with sub-normal birth weight and length often seen in people with PTPN11 and RAF1 mutations [6].

However, infants with NS may fall within normal growth parameters at birth, with their birth length, weight, and head circumference all appearing to be within a normal range [2, 3, 18, 19]. In the first year of life, growth (height and weight) decelerates rapidly, with infants presenting with failure to thrive [3, 6, 15]. During early childhood, growth may fall more than 2 standard deviation scores (SDS) below the mean [6, 10]. Puberty is often delayed, and bone age may be retarded by up to 2 years [2, 6, 10, 20]. Delayed bone ageing may indicate that there is growth potential beyond puberty, as people with NS often grow until their early twenties (commonly males) [6, 19‒21].

Due to their short stature and reduced height velocity (HV), patients often need to be treated with growth hormone (GH) [6, 22]. The efficacy and safety of GH therapy in childhood growth disorders such as NS, Turner syndrome, and GH deficiency are currently being extensively studied across the world [6, 10, 22‒28]. The clinical presentation of these disorders is heterogenous, and response to therapy varies [20, 28]. Patients with NS tend to be shorter than patients with Turner syndrome or idiopathic GH deficiency [23, 28] and experience delayed puberty, which affects treatment outcomes [2]. In addition, patients with NS pose a treatment risk due to cardiac comorbidities and higher risk of malignancy [2, 6, 7, 10, 19, 29].

GH treatment for short stature due to NS is now approved in the USA, countries of the European Union, Japan, Israel, Brazil, South Korea, Switzerland, Canada, and Argentina [30‒34]. Published reference growth charts are used to assess growth during childhood, and this determines the need for treatment. However, current use of non-NS-specific references for comparison of NS height data may present bias in results. For example, delayed puberty in NS may be incorrectly represented as growth deceleration [20].

Growth charts specific to NS were published by Ranke et al. [35] more than 30 years ago, in a 1-year study with 144 participants from two West German centres. Similar charts from a genetic standpoint were published in 1986 [36] and 2012 [18]. These charts are commonly used as NS-specific references in clinical trials; however, more current data regarding HV, comorbidities such as cardiac abnormalities, and puberty are lacking.

Data from a more recent study conducted by Isojima et al. [15] in Japan were used to construct reference growth charts for the paediatric Japanese NS population. While these survey-based data represent a wide pool of patients with NS, to accurately assess growth and NS-specific characteristics over time, there is a need for comprehensive longitudinal data [7].

The current longitudinal study followed a cohort of GH treatment-naïve patients with NS over 15 years and presents updated growth charts specifically for patients with NS in Europe. This useful tool provides a reference standard that will enable effective monitoring of growth in patients with NS and provides an accurate method of documenting response to GH therapy in future clinical trials. In addition, in-depth correlations between genotype, growth, and clinical features of this patient cohort will help clinicians to better understand the condition and possible clinical outcomes.

This study aimed to develop NS-specific growth charts and identify correlations between growth, genotype, and clinical features of patients affected by NS. The primary endpoint was to develop growth charts for patients with NS aged 0–20 years. Secondary endpoints included correlations between genotypes and the following: final height, HV, bone age, thyroid anomalies, cardiovascular anomalies, pubertal status, and age at menarche in female patients.

This was a monocentric, observational, retrospective, non-interventional, non-profit study of patients with NS. The study was conducted at “Bambino Gesù” Children’s Hospital in Rome, Italy. All patients with NS seen during 2005–2022 were eligible for inclusion. Inclusion criteria included participants with a clinical diagnosis of NS who had given informed consent (by parents/legal guardians for minors). Participants were excluded if they were currently receiving or had previously received GH treatment. If participants started receiving GH treatment after 2020, data pertaining to treatment were not recorded.

During routine visits, participants (or their parents/legal guardians) were asked to provide consent for data collection. Information was derived from medical records covering the entire period between diagnosis and last follow-up, or until the patients reached 20 years of age. Birth data were retrospectively collected from birth/hospital records. The “Bambino Gesù” Children’s Hospital’s usual clinical practice indicates that patients who are over the age of 18 years should continue to be followed up if they have a HV greater than 2 cm/year. In line with the “Bambino Gesù” Children’s Hospital guidance (continue data collection beyond age 18 years if HV is greater than 2 cm/year), data collection in most patients terminated at 18 years. Data were entered in an electronic database restricted to investigator use. For exploratory purposes, we retrospectively analysed longitudinal data (collected during routine visits at 6–12-month intervals) that were retrieved from the registry. For all study participants, auxological data collected were standing height, weight, body mass index (BMI), HV, and Tanner stage; standing height was measured with a Harpenden stadiometer.

Personal data were not collected for the purpose of the study, and all patients were anonymised with randomly generated alphanumerical IDs. Information collected included: demographic/baseline parameters: year of birth, birth weight (kg, SDS), birth length (cm, SDS), genetics (Noonan-associated mutations), parental height (cm, SDS), target height (cm, SDS), sibling(s) height (if applicable, cm, SDS), presence of thyroid anomaly (if applicable), thyroid nodules (if applicable), date of diagnosis of thyroid anomaly/nodule, presence of cardiovascular anomaly (if applicable); growth parameters: height (cm, SDS), weight (kg, SDS), HV (cm/year, SDS), final height (cm, SDS) – measurements of height/length and weight were collected at least yearly – pubertal status, and date of menarche (females). Bone age was calculated using the Greulich and Pyle method by expert readers Marco Cappa and Graziamaria Ubertini [37]. Cardiac anomalies were classified as mild, moderate, or severe, based on practice guidelines according to the respective conditions [38‒41]. Pulmonary stenosis was specifically defined as: mild PS, if right ventricular pressure (RVP) was lower than 50% of systemic pressure; moderate PS if RVP was between 50 and 74% of systemic pressure; and severe PS when RVP was >75% of systemic pressure. Pubertal status was assessed according to Marshall and Tanner staging [42, 43]. Different ages for each sex were selected for analyses (5, 9, and 13 years for females; 5, 10, 14 and 15 years for males) due to females usually reaching puberty before males [44], and bone maturation in males occurring up to 1 year later than females [45].

Descriptive data were summarised as mean and standard deviation (SD) (continuous variables), or counts and percentages (categorical variables). Between-group comparisons were performed using the Mann-Whitney U test for continuous variables and the χ² test or the Fisher’s exact test for categorical variables.

Sex-specific centile curves for height, weight, BMI, and HV were constructed. To assess the growth process of children with NS, percentile reference curves were constructed with RefCurv software [46]. The software uses, as the underlying statistical engine, the R add-on package for “Generalized Additive Models for Location Scale and Shape” (GAMLSS) developed and implemented by Rigby and Stasinopoulos [47, 48]. For statistical computations of percentile curves, the nonparametric Lambda-Mu-Sigma method by Cole and Green was adopted as a standard procedure applied in many paediatric studies on reference curves [49‒51].

Model class is defined by the Box-Cox Cole Green distribution [52] and penalised splines as smoothing for the distribution parameters L, M, and S. Model selection, based on the Bayesian information criterion (BIC) [53], was performed to find the appropriate setting of the parameters L, M, and S and to avoid overfitting. Penalised splines have a degree of freedom (df; 0–5) to be predefined by the user; the df setting of the model with the lowest BIC is considered as best for the chosen dataset; thus, df values chosen for each parameter included height – females (L0, M5, S0), males (L0, M5, S2); weight – females (L1, M5, S0), males (L0, M5, S2); BMI – females (L0, M4, S0), males (L0, M4, S0); HV – females (L0, M5, S0), males (L0, M4, S0).

Evaluation of outliers was also performed, as they might have had an adverse effect on the model fitting. Statistical results are shown in the tables and figures. Analyses were conducted with SAS software (release 9.4) and R software (version 3.5.2). Growth curves for height, weight, and BMI were constructed with reference to Cacciari et al. [54], and HV in reference to Tanner et al. [55].

Overall, 190 eligible patients with NS were identified. Study population characteristics are reported in Table 1. The median (IQR) age of participants was 14.01 (9.05–19.25) years, and 55.8% were male. Cardiovascular anomalies were observed in 85.3% of participants, most commonly PS (52.6%) and atrial septal defects (ASDs) (36.8%). Most of the cardiac conditions were mild to moderate in severity (81.1%). Almost half (48.1%) of the male patients had unilateral or bilateral cryptorchidism. The most frequently detected mutations were in PTPN11 (66.3%), followed by SOS1 (13.9%).

Sex-specific centile curves for height, weight, HV, and BMI from 0 to 18 years are shown in Figure 1a–d. Information relating to individual measurements is given in online supplementary Table S1a–h (for all online suppl. material, see https://doi.org/10.1159/000540092). Although data were collected in some individuals (7 females, 15 males) over the age of 18 years (until they reached 20 years), most patients in our study had HV less than 2 cm/year at 18 years and thus were not followed up past this age. Consequently, there were insufficient data in patients 18–20 years to create accurate growth curves within this age range.

Sex-specific growth curves relative to height, weight, and BMI are shown in Figure 2a–c; information relating to individual measurements is given in online supplementary Table S1a–h. Reference growth curves for the general Italian population were derived from Cacciari et al. [54]. For both sexes, the 50th percentile (median value, 0 SDS) of height and weight for patients with NS overlaps with the 3rd percentile (<–2 SDS) for the general Italian population; and the 97th percentile (3 SDS) of NS individuals corresponds to the 50th percentile (0 SDS) of the general Italian population (Fig. 2a and b). BMI for both females and males with NS was below the reference curve for the general Italian population. This trend was seen up to approximately 12 years of age, after which patients seemed to progressively gain weight. Both sexes crossed the 97th percentile (3 SDS) of the general Italian population around 17 years of age without reaching a plateau, as shown in Figure 2c.

Mean height and HV for both males and females with PTPN11 mutations versus “other mutations” at different ages (5, 9, and 13 years for females; 5, 10, and 14 years for males) were analysed. Females with a PTPN11 mutation had a lower height than those with “other mutations” at all ages. This was statistically significant at 5 years of age (p = 0.04), although statistical significance was not reached at 9 and 13 years (both p = 0.13). Males with a PTPN11 mutation at ages 5 and 10 years had significantly lower height than those with “other mutations” (p = 0.04 and p = 0.008, respectively). The gap in height persisted at 14 years, although statistical significance was not reached. No significant differences were observed for either sex for HV at any age. HV in males at 10 and 14 years old was non-significantly lower for those with PTPN11 mutations than in those with “other mutations.” Bone age at 5, 9 and 13 years old for females and at 5, 10 and 14 years for males was similar to actual age for individuals with a PTPN11 mutation and for those with “other mutations,” respectively; there were no significant differences. It should be noted that bone age was not recorded for all patients (online supplementary Table S2).

Information on age at menarche was available for 11 females with the PTPN11 mutation and 12 females with “other mutations” (online supplementary Table S3). Onset of menarche was similar between females with a PTPN11 mutation and those with “other mutations,” with a mean ± SD age of 13.4 ± 2.7 years in the PTPN11 group and 14.1 ± 1.6 years in the “other mutations” group (p = 0.10). Tanner staging by sex is shown in online supplementary Table S4. Females with a PTPN11 mutation were further along in their pubertal development than those with “other mutations”; this was not statistically significant. A total of 70.6% (n = 12) of females with a PTPN11 mutation had reached Tanner stage III or IV by 13 years old, compared with 66.6% (n = 6) of females with “other mutations”. At age 13, 41.2% of females with PTPN11 mutation (n = 7) were Tanner pubertal stage IV and 5.9% (n = 1) were stage V, while, among females with “other mutations,” 22.2% were stage IV (n = 2) and none were stage V (p = 0.76). Among males, at age 14 years, there were no significant differences in pubertal status (PTPN11: 46.2% [n = 6] stage III and 30.8% [n = 4] stage IV; “other mutations”: 50.0% stage III [n = 3] and 33.3% stage IV [n = 2; p = 1.0]). Mean ± SD testicular volume at age 14 was 12.5 ± 4.3 mL in the PTPN11 group (n = 13) and 14.7 ± 6.3 mL in the “other mutations” group (n = 6; p = 0.41).

Prevalence of cardiac and thyroid disease in participants with PTPN11 mutations versus those with “other mutations” is shown in online supplementary Table S5. Presence of a PTPN11 mutation was not associated with significant differences in prevalence of cardiac disease. Prevalence of thyroid disease was slightly, but not significantly, lower among patients with a PTPN11 mutation compared with those with “other mutations” (5.6 vs. 9.5%).

Anthropometric indices (height, weight, BMI, HV) at different ages according to sex and presence and severity of cardiac disease are shown in online supplementary Table S6. There were no statistically significant differences for any of the parameters by sex or age.

Anthropometric indices (height, weight, BMI, HV) at different ages according to sex and genotype (PTPN11 vs. “other mutations”) are shown in Table 2. Females with a PTPN11 mutation showed significantly lower height and weight at 5 years of age (p = 0.04 and p = 0.015, respectively) and significantly lower BMI at 9 years of age (p = 0.017) compared with those with “other mutations.” Small differences persisted at age 13 years, but this was not significant. Males with a PTPN11 mutation showed significantly lower height and weight at 5 and 10 years of age (p = 0.04 and p = 0.02 at 5 years, p = 0.008 and p = 0.008 at 10 years, respectively). A significantly lower weight and BMI (p = 0.026 and p = 0.025, respectively) at 14 years of age and significantly lower height and weight (p = 0.04 and p = 0.014, respectively) at age 15 years were observed.

This study, conducted over 15 years with 190 participants is, to our knowledge, the largest and longest longitudinal study of NS to date. We present updated long-term NS-specific growth charts to be used both in conjunction with those previously published [15, 18, 35, 36] and as the potential updated reference standard for patients with NS in Europe.

Using anthropometric data, we have produced sex-specific centile curves for height, weight, BMI, and HV, using growth charts published by Cacciari et al. [54] as a reference for the general Italian population. Median values for height and weight for both sexes with NS generally overlapped with the 3rd centile of the general population. This is consistent with previous studies describing rapid deceleration in growth in the first year of life [6, 10, 15], with height SDS and weight SDS of patients with NS shown to be between 0 and –2 throughout childhood [10, 56‒58]. Notably, at around 12 years of age, a progressive increase in weight was seen in both sexes, which did not reach the plateau typical of the general population after 16 years of age; BMI curves reflected this trend. This could be due to delayed puberty, possibly shifting onset of the typical plateau for those with NS to a later time point (after 18 years of age). Further adult analyses are necessary to distinguish whether BMI normalises after the final growth spurt or whether patients with NS continue to gain weight into adulthood.

We collected additional data on clinical features and genotypes of patients with NS, making this the most recent of only two long-term studies, the other conducted by Shaw et al. in 2007, over a 12-year period [56]. In this cohort, PTPN11 was the most common mutation, followed by SOS1. Thus, our analyses were conducted using the PTPN11 mutation as the reference for statistical comparison of anthropometry (height, weight, BMI, HV), pubertal status, and comorbid anomalies in different age groups and sexes.

Both females and males with PTPN11 mutations had lower BMIs than those with “other mutations”, and showed significantly lower height (p = 0.04 for both sexes) and weight (p = 0.02 for both sexes) at 5 years of age. Females with PTPN11 mutations continued to be shorter at 9 and 13 years of age, but this was non-significant, possibly due to the small sample size. It is interesting to note that, for males, significance was maintained at 10 years of age (p = 0.008 for both height and weight) and at 15 years of age (p = 0.04 for height; p = 0.014 for weight). This indicates that males’ growth may be more impaired in those with a PTPN11 mutation throughout the growth period. There were no significant differences in HV at any age for either sex between patients with PTPN11 mutations and those with “other mutations.” A study by Ferreira et al. showed higher HV in patients with PTPN11 mutations compared with those with other mutations before GH treatment (mean ± SD 4.3 ± 1.0 cm/year vs. 3.9 ± 1.4 cm/year); however, patients with other mutations had a better response to treatment than those with PTPN11 mutations [59]. A reason for the growth reduction may be that increased concentrations of GH, along with mild GH resistance, have been observed in patients with PTPN11 mutations, which may contribute to shorter stature and reduced growth [19, 60]. This is comparable to studies showing stunted growth in individuals with a PTPN11 mutation compared with those with other mutations [1, 15, 19, 58, 59, 61], although several studies have shown no significant difference between genotypes regarding height and weight [33, 56, 62, 63].

It is well known that patients with NS have delayed puberty, with bone age delayed by approximately 2 years [2, 6, 10]. Age at menarche for females with NS has been reported to be approximately 15 years [3, 6], whereas, in this study, average age at menarche was approximately 13.5 years. Females with PTPN11 mutations had earlier onset of menarche compared with those with “other mutations” (p = 0.10) and were further along in pubertal development at 13 years old. No statistical differences were observed for the pubertal status of males; testicular volume was marginally lower in the PTPN11 group. Pubertal development seems to be in line with the general population for both sexes, although accurate correlation of age with pubertal stage is difficult, as several studies have described varying puberty onset for children of differing population groups, with Italian females reaching menarche at 12 years [42, 43, 64‒66]. Although puberty data collected in this study represent only a small pool of patients, to our knowledge, information relating PTPN11 mutations versus “other mutations” to pubertal development has not been previously reported, which is a strength of this study. Further studies with larger sample sizes are needed to accurately evaluate pubertal development in patients with NS and the potential need for pubertal induction therapy (with or without GH treatment).

Total incidence of cryptorchidism was less than the 60–77% previously reported in other patients with NS [56, 67]. Most participants had mild-to-moderate cardiovascular anomalies, most commonly PS (52.6%) and ASDs (36.8%). In this study, 41.1% of patients with an ASD and 54.8% of patients with PS had a PTPN11 mutation, compared to 39.7% and 52.4% of patients with “other mutations,” respectively, but these were not statistically significant. However, this is comparable to the long-term study by Shaw et al. [56], which showed that, when compared with patients with other mutations, PS was present in 77% of patients with PTPN11 mutations compared with 56% of patients with other mutations, and ASDs were present in 20% compared with 13% of patients, respectively. A second study by Yoshida et al. [63] noted similar findings, with 56% of patients with a PTPN11 mutation having either PS or an ASD compared with 22% and 15% of patients with other mutations, respectively. Both studies showed HCM to be common in patients with other mutations, which would likely reflect the results shown here, as HCM was the third most common cardiac disease.

The van der Burgt criteria used for the clinical diagnosis of NS lists PS and/or HCM as major criteria [7, 57]. PS is a common cardiac anomaly in patients with NS but, in contrast to other studies [7, 57, 68], only 12.6% of patients in this study had HCM. However, it has been noted that PTPN11 and SOS1 mutations are associated with PS, whereas RAF1 and RIT1 mutations are more commonly associated with HCM [7, 8, 14, 16, 19]. This reflects the patient population studied here, as RAF1 and RIT1 mutations were identified in only 7.4% of patients. Considering the results of this study, and those of other studies listing ASD as one of the most common cardiac anomalies [15, 26, 68], ASD may need to be considered as a major clinical diagnostic criterion. Furthermore, when assessing the impact of GH therapy on cardiac disease, due to HCM being a major criterion, studies often focus on ventricular thickness [6, 69‒71], whereas it may be prudent to closely examine the atria as well.

The impact of height, weight, BMI, and HV on presence and severity of cardiac disease were analysed at different ages in both males and females; however, no significant differences were observed. It follows that the severity of cardiac disease in NS may not be impacted by growth. This is reassuring, as GH therapy may pose a risk to patients with cardiac disease, due to the relatively unknown effect of exogenous GH on the heart [6, 26, 72].

As this was a single-centre study, data were homogenous, and all variables were analysed and treated similarly, which is a study strength. A limitation of these findings regarding PTPN11 mutations is that the subgroup containing “all other mutations” is clinically heterogenous. This study was retrospective and monocentric, thus there is a need for a wider representative sample.

This longitudinal study represents the most recent and largest cohort of patients with NS followed up over 15 years. Longitudinal growth curve data, including HV, generated from this analysis will provide an important, useful tool for clinicians in the diagnosis and treatment of NS. Additionally, analyses related to genotype and clinical features of the disease provide updated and comprehensive insights into NS, allowing correlation between genotype and clinical features in future studies.

Editorial support was provided by Amy Hepple, Beverly La Ferla, and Helen Marshall, of Ashfield MedComms, supported by Novo Nordisk Health Care AG.

This retrospective review of patient data did not require ethical approval in accordance with local/national guidelines. The “Bambino Gesù” Pediatric Hospital is a Scientific Institute for Clinical Research (IRCCS) and is Joint Commission International (JCI) affiliated (accredited by the JCI in 2006). For studies using human participants, written informed consent was obtained from all adult participants, and for all participants aged under 18, written informed consent was obtained from their parent/legal guardian/next of kin, for every outpatient and inpatient evaluation.

Marco Cappa has been part of a consulting board for Novo Nordisk, as well as for Sandoz and Pfizer. Francesco d’Aniello, Maria Cristina Digilio, Maria Giulia Gagliardi, Chiara Minotti, Giusi Graziano, and Graziamaria Ubertini have no conflicts of interest to declare. Antonio Nicolucci received research support from Novo Nordisk, Italy. Alberto Pietropoli and Pier Paolo Leoncini are employees of Novo Nordisk.

Medical writing of this manuscript was funded by Novo Nordisk Health Care AG. Novo Nordisk was not involved in study design, collection, analysis, and interpretation, nor the writing of the report, nor decisions to submit the manuscript for publication other than to the extent required for those in the author group who are Novo Nordisk employees to meet authorship criteria. CORESEARCH received funding for statistical analysis from Novo Nordisk.

Marco Cappa was involved in the conceptualisation, funding acquisition, investigation, methodology, supervision, validation, visualisation, writing of the original draft, and review and editing of the manuscript. Francesco d’Aniello was involved in conceptualisation, data curation, formal analysis, methodology, project administration, software, writing of the original draft, reviewing, and editing of the manuscript. Maria Cristina Digilio was involved in the supervision of genetic data, including clinical and molecular descriptions, genotype-phenotype correlations, and review of the manuscript. Maria Giulia Gagliardi was involved in cardiological data collection, supervision, and review of the manuscript. Chiara Minotti was involved in genetic data collection and review of the manuscript. Pier Paolo Leoncini was involved in review of the manuscript. Alberto Pietropoli was involved in review of the manuscript. Antonio Nicolucci was involved in statistical analysis, interpretation of results, and critical review of the manuscript. Giusi Graziano was involved in statistical analyses performance and statistical plan writing. Graziamaria Ubertini was involved in the conceptualisation, data curation, formal analysis, funding acquisition, methodology, project administration, software, supervision, validation, writing of the original draft, reviewing, and editing of the manuscript.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author, Marco Cappa, upon reasonable request.