Neurofibromatosis–Noonan syndrome (NFNS) (OMIM 601321) is an extremely rare genetic entity described for the first time by Allanson et al in 19851 combining the clinical phenotypes of two conditions: neurofibromatosis type 1 syndrome (NF1) (OMIM 162200), characterised by a neurodermatologic presentation,2 and Noonan syndrome (NS) (OMIM 163950) distinguished by the association of severe short stature (SS), typical facial dysmorphia (TFD) and predominantly right congenital cardiopathies.3 Patients with NS have distinctive facial features, which changes with age, including: hypertelorism, low-set posteriorly rotated ears, epicanthus folds, triangular face shape, short neck and micrognathia.4

It is known that NFNS emanates from a dysregulation of the rat sarcoma oncogene homologue/mitogen-activated protein kinase (RAS/MAPK) pathway,5 a paramount cascade implicated in cell mitosis, function, differentiation and apoptosis.6 RASopathies encompass a constellation of diseases emerging from genetic alterations of different components or regulators of this latter pathway, such as NS, NF1, Costello syndrome, Legius syndrome, NS with multiple lentigines, NS with loose angen hair, cutaneous facial cardiac syndrome and NFNS.7

These diseases have in common an alteration of the RAS/MAPK pathway and overlapping features such as facial dysmorphia, congenital cardiopathies, cutaneous abnormalities, ocular and neurological impairment, as well as an increased risk of tumours.8

The underlying genetic defect of NF1 is haploinsufficiency of neurofibromin, a protein encoded by NF1 gene located on chromosome 17. This protein physiologically deactivates guanosine triphosphate) RAS into guanosine diphosphate RAS, and thus impeding the pathway. As a result, NF1 is associated with an upregulation of this latter cascade, which fosters the downstream effectors promoting tumour development.7

As for NS, it is the result of mutations affecting several genes (PTPN11, SOS1,SOS2, BRAF, RAF1, KRAS, LZTR1, MAP2K1, RIT1, NRAS, RASA2, RRAS2), controlling multiple elements of the cascade such as phosphatase and kinase and consequently, it is also tightly involved in a rampant RAS/MAPK pathway.7

Although merging clinical features of both NF1 and NS, the genetic background of NFNS is further from being a mere occurrence of mutations affecting the genes mentioned above. It is actually a matter of debate and still a fertile ground for research.

Given the paucity of reported cases and the absence of comprehensive reviews on NFNS, gaps remain in our understanding of this genetic condition. We aim through this systematic review to shed light on the broad clinical spectrum, the epidemiological features, the underlying genetic defects and the associated comorbidities of NFNS.

GA at birth was available in 22 cases. For 13/22 patients (59.1%), a normal term delivery was found and in 40.9% of cases, patients were born preterm (GA<37 weeks). The mean GA was 35.3 weeks (min, max: 28, 40 weeks) and the mean weight at birth (n=19) was 2468 g (ranging from 1157 to 4000 g), corresponding to a mean percentile of 52.5. Both fetal hypotrophy (weight at birth<10th percentile) and large for GA (weight at birth>90th percentile) were diagnosed in 2/19: 10.5% of patients. Mean height (n=10) and HC (n=6) were 47.2 cm (40.7 percentile) and 33.5 cm (60.9 percentile), respectively. Macrocephaly (HC at birth>90th) was encountered in one case (1/6: 16.7%) and small for gestation age (height at birth<10th) in 2/10: 20% of patients.

At diagnosis, 34/45: 75.6% of patients had SS, particularly 28/45: 62.2% had severe growth retardation.

The clinical characteristics of NS and NF1 were not detailed in two patients and in one sole case, respectively.

Café au lait (CAL) spots were the hallmark of NFNS, present in the majority of cases 93/94: 98.9%. The other symptoms related to NF1 were: NF (cutaneous or subcutaneous) 52/94: 55.3%, lentigines (axillar or inguinal) 47/94: 50%, Lisch nodules (LN) 33/94: 35.1%, unidentified bright objects on MRI 12/94: 12.7%, plexiform NF 7/94: 7.4%, cerebral glioma: 5.3%, hamartomas 3/94: 3.2%, seizures 3/94: 3.2% and osteoporosis 1/94: 1%.

As for clinical signs of NS, they were dominated by: TFD 88/93: 94.6%, pectus abnormalities (PA) 59/93: 63.4%, broad neck (BN) 59/95: 63.4% and followed by mental retardation (MR): 18/93: 19.3%, cubitus valgus 19/93: 20.4%, wide spreading nipples and broad thorax in 12/93: 12.9%, each. Additionally, hyperextensible joints were seen in 6/93: 6.4% and hypotonia in 9/93: 9.7%. Cryptorchidism was a frequent finding in the setting of NFNS, identified in 12/49: 24.4% of cases.

Macrocephaly, scoliosis and cardiopathies, common traits in NS and in NF1, occurred in 24/92: 26%, 39/92: 42.4% and 34/92: 36.9%, respectively. 15 patients had learning difficulties, corresponding to 16.3% (n=92). Online supplemental annex 9 describes the different types of cardiopathies found across the studies.

Associated features encompassed tumours (other than neurofibromas and LN) in 17/92: 18.5% (as illustrated in online supplemental annex 10), autoimmune diseases 4/92: 4.3%, genitourinary conditions 9/92: 9.8%, ophthalmological aberrations (other than LN) 16/92: 17.4% and otorhinolaryngologic abnormalities 5/92: 5.4%.

Regarding endocrinopathies, they were observed in 11/92: 11.9%, prevailed over by isolated growth hormone deficiency (GHD) in three cases. It was imputable to pituitary hypoplasia in one individual, whereas pituitary MRI was normal in one of the remaining cases and not stated in the other. As for hypogonadotropic hypogonadism, it was attributable to hydrocephalus in one case and to chronic severe cardiopathy in the other patient (online supplemental annex 11).

Genetic analysis of the NF1 gene was performed in 72 cases (75.8%) and revealed a genetic defect in 63: 87.5% while biomolecular probing for mutations related to NS genes was carried out in 52 patients (54.7 %) and unravelled an anomaly in 5: 9.6%. PTPN11 gene was the main gene investigated in the setting of NS (42/52: 80.8%).

Concomitant mutations of NF1 gene and genes related to NS were found in four cases (4/52: 7.7%). In one case, a mutation of the PTPN11 gene was detected but not in the NF1 gene (1/52: 1.9%).

The most frequent mutation types in NF1 gene were deletion and missense. Figure 2 is a representation of the genetic mutations spanning different exons of the NF1 gene. It is noteworthy that in many cases, the exact location of the mutation was not determined, and thus it was not represented in figure 2.

Mutation sites throughout the NF1 gene. Frequency of mutation sites throughout the NF1 gene.

A wide array of biomolecular analytic tools was employed across the included studies (online supplemental annex 12).

NFNS is a very rare condition, with enigmatic, blurred features. To date, this is the first attempt to systematically assess and comprehend its clinical characteristics as well as its genetic background.

NFNS is a distinct entity among the group of RASopathies. Src-homology-2-containing protein tyrosine phosphatase 2 and neurofibromin encoded, respectively, by PTPN11 and NF1 are functionally related in the common RAS/MAPK signal pathway (figure 3).51

Simplified scheme of the RAS/MAPK pathway. Different ligands, such as growth hormone or insulin, activate the tyrosine kinase receptor and trans-autophosphorylate. Docking proteins such as growth factor receptor bound protein 2 (GRB2) are consequently phosphorylated, allowing them to interact with sequence homology (SH) molecules such as Src-homology-2-containing protein tyrosine phosphatase 2 (SHP2), capable of recruiting Son Of Sevenless1 (SOS1), which enables the transformation of guanosine diphosphate (GDP) to guanosine triphosphate (GTP). Conversely, neurofibromin does the exact opposite. Once rat sarcoma oncogene homologue (RAS) protein activated, it binds to rapidly accelerated fibrosarcoma (RAF) in cell membrane. Following RAF activation, a cascade of protein phosphorylation occurs such as mitogen-activated extracellular signal-regulated kinase (MEK) and extracellular signal-regulated kinase (ERK). These latter effectors prompt downstream nuclear and cytosolic factors to stimulate cell proliferation, differentiation and survival.51 MAPK, mitogen-activated protein kinase

Our present study unravelled very interesting results. Family occurrence of NFNS was depicted in 10 out of 69 included families (14.5%), underscoring the need to screen and counsel affected families.

Common clinical features could distinguish NFNS. At birth, patients with NFNS had frequently premature birth (40.9%), a prevalence much higher than the global prematurity frequency (9.9%) as stated by a recent survey.52 Prematurity is found in 58% of RASopathies in general,53 occurring in 80% of mothers with NS according to a small retrospective study54 and 30.8% of pregnant women with NF1.55 It is actually known that RAS/MAPK pathway has a major role in the normal development of placenta, thus explaining partially this finding.56 Macrocephaly is another potential mechanism of prematurity in NFNS, since it exposes to obstetric complications.57 In our research, macrocephaly was present in 16.7% of patients, a percentage much higher than the international prevalence of 9%.58 However, these results are congruent with a study that compared of 1410 individuals with NF1 syndrome and their matching controls, concluding to a larger HC in NF1.59

At diagnosis, hallmark features of NFNS were CAL spots, TFD of NS, postnatal SS, PA, BN and lentigines.

The occurrence of CAL lesions in NF1 is attributable to marked secretion of hepatocyte growth factor and stem cell factor by fibroblasts in the dermis resulting in the overexpression of melanocytes and consequently skin pigmentation. A similar mechanism could be implicated in NFNS.60

A major downstream effector of the RAS/MAPK pathway is extracellular signal-regulated kinase. It is phosphorylated, expressed in the neural crest and is involved in the early forming structures of the ear, eye and heart, in murine models.7 A comparable process in humans can explain facial dysmorphia and congenital heart diseases in patients with NFNS.

Postnatal SS is a common feature of RASopathies, imputed probably to growth hormone insensitivity through a post receptor defect,61 thus explaining its frequency in NFNS. Nevertheless, isolated GHD was present in 3.3% of cases. Nutritional causes could also be incriminated, since some patients exhibited feeding problems. Definitely, more research is warranted to explore the growth hormone axis in NFNS.

Mean age at diagnosis of NFNS was 14.7 years, suspected on account predominantly of family investigation, neurofibromas, facial dysmorphia and SS. For both NF1 and NS, the mean age of diagnosis was 8 years.62 63 A substantial disparity of diagnostic age is noted across studies, some patients were diagnosed at birth64 while others as late as 69 years.28 This age gap in NFNS could be ascribed to the broad spectrum of its clinical presentation, making it challenging to diagnose in mild cases, along with a lack of knowledge of this syndrome.

Attempting to grasp the relation between NFNS, NF12 63 and NS,3 63 65–70 we confronted main characteristics of these conditions, as outlined in table 2. While we are aware of the statistical inaccuracy of this comparison, we thought it may help delineate NFNS’s peculiarities.

Confronting characteristics of NFNS, NF1 and NS

Accordingly, a combined effect of NF1 and NS seems to result in NFNS’s phenotype, particularly: SS, MR, cryptorchidism, macrocephaly, otorhinolaryngeal abnormalities and tumour development. However, some peculiarities of NFNS are noteworthy: the relative high frequency of scoliosis and rarer occurrence of plexiform NF as well as ophthalmological anomalies.

An increased frequency of tumorigenesis is noted in our systematic review. Other RASopathies share this feature with NFNS, and multiple studies highlighted this aspect.7 56 Somatic mutations of RAS proteins are detected in 20% of patients with malignancies, asserting its critical role. An inherent trait of the RAS/MAPK pathway is its pro-oncogenic effect, mediated essentially by continuous stimulation of cell proliferation, angiogenesis, resistance to apoptosis and genomic instability, thereby the susceptibility to malignancies.6

The molecular determinants of NFNS have been for long time a matter of debate. Some authors suggested that NFNS is a variant of NF115 or NS,40 whereas others stipulated that it is a unique syndrome.1 Other researchers hypothesised that NFNS is a random occurrence of both conditions.46

The theory of co-occurrence of genetic mutations of both NF1 gene and NS-related genes (PTPN11, SOS1, SOS2, BRAF, RAF1, KRAS…) is very unlikely. In fact, we would expect to find mutations of PTPN11 in about 50% of patients with NFNS, since it is the major underlying genetic cause of NS,70 which was not the case. Still, these results must be interpreted with caution as the molecular analysis of NS genes was carried out only in 54.7% of cases. Additionally, PTPN11 was the main gene examined in 80.8% of patients, which underestimates genetic diagnosis of NS (online supplemental annex 8).

Otherwise, Colley et al reported the presence of NS features in 9/94 individuals with NF1 (9.4%), which dismisses the possibility of a chance occurrence of both conditions.71

Subsequently, the most plausible theory seems to be that NFNS is merely a subset of NF1, which is endorsed by the detection of NF1 mutations in 87.5% of cases, although not all with confirmed pathogenicity. The logical question that arises from this statement is: how come not all patients with NFNS have NF1 mutations? Two possible responses: first the possibility of a mosaic disease limiting accurate genetic diagnosis72 and the second is the use of genetic methods and techniques exploring partially the NF1 gene. Actually, It is recommended to employ a combination of different techniques encompassing fluorescent in situ hybridisation, Multiplex Ligation-dependent Probe Amplification, direct sequencing etc73 and more recently, novel molecular techniques for instance next-generation sequencing, capable of detecting mutations of the NF1 gene in more than 98% of cases,74 which was not employed in most of the included studies.

One logical inquiry arises from this latter hypothesis: why is it some patients with NF1 mutations develop NS features while others do not? Epigenetics may be the answer in this case.

Yapijakis et al described a family where the children had a phenotype resembling NFNS: SS, and suggestive facial dysmorphia of NS associated with CAL macules and cutaneous NF. The probands and their father had a nonsense mutation of the NF1 gene. Intriguingly, during both pregnancies their mother took hydantoin, which resulted in a fetal hydantoin syndrome, mirroring many aspects of NS, but still distinguished by hypoplastic nails. Thus, the authors suggested that NFNS may be the conjunction of genetic causes, NF1 gene mutation for instance and epigenetic factors (hydantoin intake).5 Nonetheless, it should be emphasised that the probands did not meet the criteria of NS, as opposed to the included patients in our study.

The recognition of NFNS as a distinct clinical entity, despite common genetic background with NF1, is paramount due to some significant differences between the two conditions. Notably, NFNS is characterised by a higher prevalence of specific comorbidities, including cardiopathies, intellectual disability, and cryptorchidism (table 2). These differences have direct implications for patient management, as they require tailored monitoring protocols and therapeutic approaches. Understanding these distinctions enhances our ability to optimise patient care and improve outcomes for individuals affected by NFNS.

While our analysis yielded interesting results and provided for the first time a comprehensive review on NFNS, it had some limitations. Most of the retrieved data derive from case reports and case series, which restricts the accuracy of our findings. In fact, only 33.3% of included studies were classified as good or excellent, which can limit the strength and reliability of our results.

Additionally, the inclusion criteria could have underestimated the true frequency of NFNS and its features, since many symptoms related especially to NF1 develop with age and patients may be paucisymptomatic initially.

Finally, a great heterogeneity of the methodology used for genetic testing across studies could influence our conclusions concerning the genetic background of NFNS.