LEOPARD syndrome (LS), or Noonan syndrome (NS) with multiple lentigines (OMIM 151100), is an autosomal dominant disorder characterized by multiple lentigines or café au lait spots, electrocardiographic abnormalities, ocular hypertelorism, pulmonary valve stenosis or hypertrophic cardiomyopathy, genital abnormalities, constitutional growth delay, and deafness.1 LS shares many features with NS (OMIM 163950), which is characterized by an association with congenital heart disease, short stature, and craniofacial malformations2 but does not usually include multiple lentigines and deafness among its manifestations. Mutations have been identified in the PTPN11 gene in 50% of NS3–5 cases and in 85% of LS cases.6,7 Mutations in the RAF18 and BRAF9 genes have also been identified in LS. Although there are no exact data on its prevalence, it is thought that NS is present in 1 in 1000 to 2500 live births, suggesting that it is an underdiagnosed disorder.10 LS is less frequent, although precise data on its prevalence at birth are also unavailable. To date, at least 200 cases have been reported in the literature and a comprehensive review of the disorder has been published recently.11 Both disorders show considerable phenotypic variability, making them difficult to identify and diagnose correctly. Genetic studies can therefore provide a useful contribution to differential diagnosis. Diagnosis is also made difficult by the evolving and changeable nature of many of the condition's features. For that reason, any objectifiable congenital heart disease can provide valuable help in identifying the syndrome. We present a phenotypic description of a series of patients with LS characterized in genetic studies of the PTPN11, RAF1, and BRAF genes, and compare them with a large series of patients with other neurocardiofaciocutaneous syndromes (NCFCS) who were also characterized by the genetic study of mutations in the PTPN11, SOS1, RAF1, BRAF, and HRAS genes.

The study included 19 patients (13 men, 6 women; age at diagnosis, 11 months to 49 years; mean age, 7.4 years). Cases were referred from 9 hospitals in 5 Spanish regions and 1 hospital in Belgrade (Serbia). Five patients were identified as familial cases (families A and B; 26% of cases; 95% confidence interval [95%CI], 9.1%-51.2%), and the rest were considered sporadic. Table 1 summarizes the phenotypic and molecular findings. Of the total, 15 met Voron's clinical criteria and 4 were considered to have a “partial form of LS”. The latter had clinical features typical of LS but did not have multiple lentigine syndrome or a family history of LS. All four were carriers of the p.Thr468Met mutation in PTPN11, which is typical of LS, and were of pediatric age at the time of the evaluation. The cutaneous features of the syndrome may therefore appear over time as the patients develop. Craniofacial (100% when typical and suggestive phenotypes were included; 95%CI, 82.4%-100%) and cardiac (88%; 95%CI, 63.6%-98.5%) anomalies, and skin lesions (84%; 95%CI, 60.4%-96.6%) were the most frequently reported characteristics, followed by short stature (37%; 95%CI, 16.3%-61.6%) and genital abnormalities (cryptorchidism in 3 patients, representing 16% of the total and 23% of men, 95%CI, 5%-53.8%). Three patients were diagnosed with hearing loss, but a sensorineural component was only observed in 1 individual (5%; 95%CI, 0.1%-26%); 7 patients had psychomotor retardation (37%; 95%CI, 16.3%-61.6%), although it was only mild in all cases. The 15 patients with multiple lentigines had a complete phenotype (79%; 95%CI, 54.4%-93.9%) and 11 of them also had café-au-lait spots (58%; 95%CI, 33.5%-79.7%). One patient with a partial phenotype also had coffee stains, while the other 3, aged between 10 months and 14 years, had no skin lesions. Two unrelated patients (patient numbers 1 and 8, with mutations p.Thr468Met and p.Gln510Arg, respectively) showed no skin manifestations at the initial assessment but developed the characteristic multiple lentigines during follow-up, which highlights the condition's evolving nature.

The most frequent congenital heart disease was hypertrophic cardiomyopathy (71%; 95%CI, 44%-89.7%), followed by pulmonary valve stenosis (35%; 95%CI, 14.2%-61.7%). Three patients (18%; 95%CI, 3.8%-43.4%) had both disorders. The only other congenital heart disease observed was aortic coarctation in 1 patient who also had hypertrophic cardiomyopathy. Coarctation in this patient resolved without corrective surgery, which would challenge the diagnosis of hypertrophic cardiomyopathy, although cardiomyopathy persisted several years after coarctation resolved. Only 2 (12%; 95%CI, 1.5%-36.4%) of the patients tested had regular cardiac testing. One patient was diagnosed at 2 months of age with Wolf-Parkinson-White syndrome following an episode of supraventricular tachycardia. One other patient also suffered supraventricular tachycardias. No patient experienced sudden death or required defibrillator implantation.

Molecular analysis of the PTPN11, RAF1, and BRAF genes detected mutations in 19 patients. The most frequent mutation was p.Thr468Met (10 patients, 53%; 95%CI 28.9%-75.6%). The p.Tyr279Cys amino acid change was identified in 2 related patients (mother and son, family B) and 3 unrelated patients (26% of the total series; 95%CI, 9.2%-51.2%). The p.Ser257Leu variant in RAF1 was detected in 2 other unrelated patients. Patients with mutations in RAF1 showed hypertrophic cardiomyopathy and some of the smallest statures in the series. Finally, we detected a change in the BRAF gene sequence in 1 patient.

To compare patients with LS and patients with other NCFCS, we included 87 patients with NS (PTPN11 mutation, 68 patients, 78%; SOS1 mutation, n=15, 17%; RAF1 mutation, n=4, 5%), 5 with cardiofaciocutaneous syndrome with a BRAF mutation, and 1 patient with Costello syndrome due to a mutation in HRAS. These patients came from 31 hospitals in 11 Spanish regions, 1 hospital in Belgrade (Serbia), and 1 hospital in Buenos Aires (Argentina). Table 2 shows the clinical features of each of these patient groups, and as compared to the LS patients. LS is related to the presence of multiple lentigines and café au lait spots (P<.001). Similarly, there is evidence of a strong association between LS and hypertrophic cardiomyopathy and greater height (P<.05).

As some authors have suggested, a classification of cardiomyopathies that takes into account their underlying molecular characteristics could contribute to more appropriate management.19 In the case of hypertrophic cardiomyopathy, the relevance of both NS and LS should be borne in mind.20 This is particularly true when, from a cardiological perspective, the accompanying phenotype can vary from fatal cases to completely asymptomatic patients. Only a high level of suspicion will allow the clinician to consider this diagnostic option when there is little evidence.

In recent years, advances in molecular biology have allowed us to elucidate many aspects of the etiology of NS and other entities with overlapping phenotypes (LS, Costello syndrome, cardiofaciocutaneous syndrome, neurofibromatosis type 1, and Legius syndrome). Some authors group these disorders under the name of NCFCS,21 RAS-MAPK syndromes, or rasopathies, as it appears that Ras/MAPK pathway dysregulation lies at the root of all of them (Figure). In NS, PTPN11 mutations have been identified in approximately 50% of patients, and mutations have been described in other genes of the RAS-MAPK pathway (SOS1,22RAF1,8KRAS,23MAP2K1,24BRAF,9NRAS,25 and SHOC226). Even today, there is no defined genetic cause in about 30% of patients diagnosed with NS.

Figure.

Ras-MAPK cascade. The binding of a growth factor to a tyrosine kinase receptor activates intracellular effectors such as SHP2, which in turn recruit guanine nucleotide exchange factors such as SOS1 that promote GDP/GTP exchange in RAS proteins. The latter are activated by phosphorylation. Consequently, Ras-GTP activates the different RAF isoforms (RAF1, BRAF), MEK (MEK1, MEK2) and, finally, ERK.

(0.2MB).

In LS, about 85% of patients have mutations in PTPN11. Eleven mutations have been described in PTPN11 but, of those, 65% correspond to 2 highly recurrent mutations (p.Thr468Met7 and p.Tyr279Cys). In approximately one third of PTPN11 negative patients, mutations have been identified in RAF1,8 another gene in the RAS-MAPK cascade. Finally, mutations have been reported in BRAF,9,27 a gene whose mutations were primarily associated with cardiofaciocutaneous syndrome.

In recent decades, molecular studies have proven useful in providing differential diagnoses between these overlapping entities, as illustrated by several cases in our series. Among our patients, p.Thr468Met was the most common mutation in PTPN11, followed by p.Tyr279Cys. During evolutionary monitoring of patients with the p.Thr468Met mutation, the appearance of multiple lentigines led to a modification of the diagnosis in 2 cases. For that reason, and following previous authors,7 we considered pediatric patients with NS due to p.Thr468Met to be a partial phenotype of LS. Mutations with changes in amino-acid Gln510 have been associated with both NS and LS28; in our series, there were patients diagnosed with NS who showed sequence changes that predicted amino acid changes due to other residues (p.Gln510Glu, p.Gln510Pro). As these were adult patients, no change in diagnosis was expected and they were not included in the study. Patient 8, on the other hand, had the p.Gln510Arg variant and developed lentiginosis during pediatric clinical monitoring, a development which led to a modification of the diagnosis. In the case of RAF1, the p.Ser257Leu mutation was identified in 2 cases. In one patient who was negative for PTPN11 and RAF1, the BRAF mutation was identified, a finding which shows the relevance of studying this gene in patients with LS. Our study confirms the substantial benefits of carrying out genetic studies in patients with LS, and illustrates its close association with the PTPN11 gene and, to a lesser extent, RAF1 and BRAF.

The main diagnostic features of this syndrome are its cutaneous manifestations, which sometimes do not fully develop until after puberty29,30; deafness, which is not as common as in the initial descriptions of the syndrome; and congenital heart disease. Although pulmonary valve stenosis was initially described as the most common heart disease observed, hypertrophic cardiomyopathy is now more frequently identified.7,11,31 In our series, this congenital heart condition was more significantly associated with LS than were other NCFCS. In our setting, congenital heart disease has led to the definitive diagnosis in a large proportion of cases, and remains the most dangerous comorbidity in these patients.32 Pulmonary valve stenosis was identified in 6 of our 19 patients, making it the second most common cardiac diagnosis. Three patients had both conditions simultaneously. Both diagnoses should lead us to consider the possibility of a broader malformation. The 3-year-old patient in whom no congenital heart disease was found remains under observation in order to rule out the evolutionary occurrence of structural heart abnormalities, as described previously.33 Although in our series there were no fatal events and defibrillator use was not required in any of the patients, the oldest in the series was referred to a cardiac unit for further study because of syncope. In fact, it has been reported that LS patients with hypertrophic cardiomyopathy appear to be at increased risk of adverse events during follow-up.32 Among the cardiac disorders, although conduction defects are frequent, the Wolf-Parkinson-White syndrome described in one of our patients is not a common manifestation.

The phenotypic overlap between LS, NS, and neurofibromatosis type 1 makes it difficult to correctly classify these patients. Noonan syndrome-neurofibromatosis syndrome, which is related to mutations in the NF1 gene34 and characterized by the association between neurofibromatosis type 1 and NS manifestations, presents characteristics that are very similar to LS. The diagnosis of a congenital heart condition such as hypertrophic cardiomyopathy in patients with Noonan syndrome-neurofibromatosis makes it advisable to start with a genetic study of PTPN11 rather than with NF1.35 Multiple lentigines and café au lait spots are common in all 3 entities. Just as café au lait spots increase in number and size with age in these entities, multiple lentigines are rare in younger children. For this reason, some authors recommend evaluating for LS when diagnosing a patient younger than 1 year of age with NS associated with café au lait spots.29 Distinguishing between multiple lentigines and nevi may be particularly complex in some patients, and some authors propose including melanocytic lesions in the LS spectrum.36 In our series, the diagnosis was limited, whenever possible, to patients with lentiginous lesions. This is illustrated by a case described in Ezquieta et al.13 The patient suffered NS stemming from the p.Asn308Asp amino acid change in PTPN11 and had multiple hyperpigmented lesions, as did his mother, who had tested negative in the mutation study. These results indicated that the multiple nevus lesions observed in mother and child were due to a cause other than mutations in PTPN11. As indicated in earlier observations, we have demonstrated that LS is associated with a taller stature than other NCFCS.

Pulmonary valve stenosis and particularly hypertrophic cardiomyopathy are cardiological findings which may provide the key to identifying the syndrome. Given the potential clinical implications of hypertrophic cardiomyopathy, patients within the clinical spectrum of NS and particularly those with LS should be actively evaluated for this disease. The criteria originally described by Voron et al. in 1976 are still helpful in classifying these patients in terms of diagnosis. Nevertheless, the carriers of LS-associated anomalies that do not meet these criteria should be monitored over time because some skin manifestations, particularly multiple lentigines, may be delayed until late puberty.

Fondo de Investigaciones Sanitarias (PI 06/1179).

None declared.