|Home | About | Journals | Submit | Contact Us | Français|
Noonan syndrome (NS) is an autosomal dominant genetic disorder associated with highly variable features, including heart disease, short stature, minor facial anomalies and learning disabilities. Recent gene discoveries have laid the groundwork for exploring whether variability in the NS phenotype is related to differences at the genetic level. Here we examine the influence of both genotype and non-genotypic factors on cognitive functioning. Data are presented from 65 individuals with Noonan syndrome (ages 4 to 18) who were evaluated using standardized measures of intellectual functioning. The cohort included 33 individuals with PTPN11 mutations, 6 individuals with SOS1 mutations, 1 individual with a BRAF mutation, and 25 participants with negative, incomplete or no genetic testing. Results indicate that genotype differences may account for some of the variation in cognitive ability in NS. Whereas cognitive impairments were common among individuals with PTPN11 mutations and those with unknown mutations, all of the individuals with SOS1 mutations exhibited verbal and nonverbal cognitive skills in the average range or higher. Participants with N308D and N308S mutations in PTPN11 also demonstrated no (or mild) cognitive delays. Additional influences such as hearing loss, motor dexterity and parental education levels accounted for significant variability in cognitive outcomes. Severity of cardiac disease was not related to cognitive functioning. Our results suggest that some NS-causing mutations have a more marked impact on cognitive skills than others.
Noonan syndrome (NS) is a multiple congenital anomaly syndrome characterized by short stature, facial anomalies, heart disease and learning disabilities. Incidence is estimated to be between 1:1000 and 1:2500 live births (Mendez et al., 1985). Germline gain-of-function mutations in several RAS-MAP Kinase pathway genes have recently been found to cause NS. Missense mutations in the PTPN11 gene are the most common cause of NS and account for approximately 50% of cases (Tartaglia et al., 2001). Mutations in SOS1, RAF1, and KRAS genes account for an additional 10–15%, 3–17%, and <5% of cases, respectively (Pandit et al., 2007; Razzaque et al., 2007; Roberts et al., 2007; Schubbert et al., 2006; Tartaglia et al., 2007). Although typically associated with cardiofaciocutaneous (CFC) syndrome, recent studies have found that BRAF mutations can also result in a NS phenotype (Nystrom et al., 2008; Razzaque et al., 2007). The genetic etiology remains unknown in roughly 30% of NS patients.
Learning disabilities are commonly cited as a key characteristic of NS, yet the causes of these impairments are poorly understood. Average IQ scores among affected individuals are lower than expected based on normative data (Lee et al., 2005). However, cognitive abilities of NS patients may range from moderate mental retardation to superior abilities. Van der Burgt and colleagues (1999) noted that individuals who displayed more severe physical features of NS performed more poorly on some cognitive tests than those with moderate features. One possible explanation for this finding is that one or more of the medical sequelae of NS could interfere with cognitive functioning. For example, research indicates that children with severe congenital heart disease tend to exhibit overall lower cognitive abilities than those with less severe heart disease (Karsdorp et al., 2006). Congenital heart defects are a primary characteristic of NS and are present in roughly 85% of patients with PTPN11 mutations (Sznajer et al., 2007). Hearing loss and motor incoordination are two additional features commonly seen in NS (Lee et al., 2005; Qiu et al., 1998). These medical characteristics could affect cognitive development in NS, but their influence has not been explored to date.
An alternative explanation for the association between severity of NS expression and cognitive ability is that certain NS mutations may have a generally more deleterious effect, resulting in abnormal physical and mental development. The signal transduction pathway in which the known NS genes act, RAS-MAP Kinase, plays a role in numerous biological processes including embryologic development (Schubbert et al., 2007). Some research suggests that dysregulation of this pathway can affect brain development. Altered activation of protein tyrosine phosphatase SHP-2, the PTPN11 gene product, can interfere with neural cell-fate decisions (Gauthier et al., 2007). Whether the molecular changes resulting from other NS mutations have similar effects on central nervous system development is not currently known.
The purpose of the present study was to examine whether some of the variability in cognitive functioning in NS could be explained by genotype, and to explore additional medical, developmental and environmental influences on these skills.
65 individuals with NS completed this study. Participants were part of a larger investigation of behavior and learning in individuals with Noonan syndrome. Families were recruited through clinics at Children’s Hospitals and Clinics of Minnesota (n=20), Children’s Hospital Boston (n=25), the Waisman Center at the University of Wisconsin-Madison (n=7), and at the annual meeting of the Noonan Syndrome Support Group (n=13). The study was approved by the Internal Review Board (IRB) at each of the participating institutions. Participants and their primary caregivers signed written informed consents prior to enrollment in the study.
Participants were recruited for the study if they had received a clinical diagnosis of Noonan syndrome from a clinical geneticist. Relevant medical and genetic information was obtained from hospital case notes requested from the child’s primary geneticist or cardiologist using HIPAA authorizations signed by the families. Review of these records was used to determine whether participants fit inclusion criteria for the study. Criteria for inclusion, based on a scoring system developed by van der Burgt et al. (1994), were identical to those used in previous studies (Roberts et al., 2007).
Intellectual abilities were evaluated using the Differential Ability Scales (DAS). This measure includes norms for individuals aged 2 ½ –18 years and provides a verbal and nonverbal cluster for all ages (Elliott, 1990). In older children (> 6 years), the Special Nonverbal Composite, which is used in our analyses as the main measure of nonverbal ability, can be further divided into spatial and nonverbal reasoning scales. All assessments were administered in a quiet room by the same examiner (EIP).
A pure-tone hearing screening was performed at the time of the assessment using a portable Beltone audiometer. Pass/fail data were collected for both ears at 20 dB for frequencies of 1000-, 2000- and 4000-Hz. A screening score (ranging from 0 to 6) was assigned based on the number of frequencies in which the participant was able to detect the tone.
Manual motor dexterity was evaluated using the Purdue Pegboard Test (Tiffen, 1968). This task requires the examinee to place small metal pegs into a series of slots as quickly as possible during a limited period of time (30 sec). Three conditions were administered: preferred hand, non-preferred hand and both hand conditions. A composite score was obtained by averaging participants’ standard scores for the three conditions. Standard scores for each trial were calculated using appropriate age norms (Gardner & Broman, 1979; Yeudall et al., 1986).
Based on a review of each participant’s medical record by a pediatric cardiologist (MEP), individuals were assigned a rating of medical severity of cardiac disease. The score was based on the Cardiologist’s Perception of Medical Severity (CSEV) scale (Demaso et al., 1991). This scale indexes cardiac severity as follows: (1) no or insignificant disorder – disorder has no impact on child’s health; (2) mild disorder – lesion requires no operative intervention, only long-term follow-up (e.g., small ventricular septal defect); (3) moderate disorder – child is asymptomatic, but has had or will require operation, easy repair (e.g., atrial septal defect); (4) marked disorder – child quite symptomatic, has had or will require major difficult repair (e.g., tetralogy of Fallot, transposition of great arteries); (5) severe disorder – uncorrectable cardiac lesions or only complex palliative repair possible (e.g., pulmonary vascular obstruction, Fontan repair, valve replacement).
Parents were asked to report their highest level of formal education. The average of the paternal and maternal years of education was calculated to index socioeconomic status. Because parental education levels have been shown to have significant impact on intellectual development, especially verbal IQ (Rowe et al., 1999), this measure was included as an additional factor in our analyses.
Gene testing reports were available for 53 participants. The remaining individuals in the sample (12 participants) had not completed any genetic testing. Of the individuals tested, 33 (62%) tested positive for a PTPN11 gene mutation and 6 (11%) tested positive for an SOS1 mutation. 13 patients tested negative for mutations in PTPN11; of these participants, eleven had not completed testing for the remaining NS genes, and two had undergone SOS1 and KRAS testing with negative results. One participant tested positive for the BRAF mutation. Although BRAF mutations are typically associated with CFC syndrome, this individual met diagnostic criteria for the NS phenotype and therefore was not excluded from the study. Specific information about the genotypes of participants is included in Table I.
The cohort included 35 males and 30 females between the ages of 4 and 18 years (M = 10.0, SD = 4.1). Family characteristics and developmental history were obtained through a review of medical reports and parent accounts. For 54 of the individuals assessed, the parents were married and living together. The remaining 11 participants were from single-parent families or were living with one biological parent and one step parent. Parental education levels ranged from some high school to advanced graduate degrees (M = 15.6 years, SD = 1.9). The cohort included one set of monozygotic twins, two families with two affected siblings, and one family with three affected siblings. In 13 patients (including the 9 participants with affected siblings in the cohort), the NS mutation was known to be inherited from an affected parent. In 4 additional participants, a diagnosis of NS was suspected but not confirmed in at least one other first degree family member. In the remaining 48 cases, NS was thought to be sporadic (non-familial). 37 of the participants (57%) had received a cognitive, learning or behavioral disability diagnosis at some point in development. The most common specific diagnoses in the cohort were Attention Deficit/Hyperactivity Disorder (29%), Reading Disability (11%), speech/language impairment (9%), Math Disability (9%) mental retardation (8%) and autism spectrum disorders (8%). One child (aged 4 years, 6 months) was nonverbal at the time of assessment.
Intellectual skills varied widely among participants, ranging from Very Low (> 2 standard deviations below the mean) to High (> 1.5 SD above the mean) levels of functioning. As a group, our NS cohort scored significantly lower on the DAS than expected based on normative data (mean = 100, SD = 15), t (64) = −6.05, p < .001. Mean scores for males and females were not significantly different, t(63) = .28, p = .78. Performance on the DAS was not significantly correlated with the chronological age of participants (r = −.02, p = .90). The distribution of scores among NS patients spanned a wide range but was shifted downward compared with the normative population (Figure 1). The group mean of 86.2 (SD = 18.4; range: 44–123) was approximately one standard deviation below the general population average.
Individuals without an established cognitive or learning disability diagnosis at the time of assessment scored significantly higher on the full scale assessment than those with an established diagnosis, t(63) = 2.16, p < .05. However, learning/cognitive disabilities may be somewhat under-identified in this population. Eight of the 23 individuals who scored in the “Low” or “Very Low” range on the DAS (> 1.5 SD below the mean) had not previously been diagnosed as learning disabled. Eleven participants in the cohort (17%) obtained a score below 70, in the range of mental retardation. This rate in NS is higher than the incidence within the general population (2%), χ2 = 73.9, df = 1, p < .001. In order to examine the possibility of ascertainment bias based on site of recruitment, we compared performance for groups tested at each of our four research sites. Cognitive scores did not vary significantly as a function of the recruitment/testing location, F(3,61) = .85, p = .47. This suggests that a similar range of abilities was seen in patients identified through medical clinics, research studies, and through the Noonan Syndrome Support Group.
Patterns of discrepancy across domains of intellectual skill were also examined. On average, verbal skills were significantly higher than nonverbal skills, t(64) = 2.84, p < .01. In order to determine the direction of discrepancies for individuals in the sample, differences between verbal and nonverbal abilities were compared with critical values for statistical significance at the p=.05 level. 14 participants (22%) had a verbal score that was significantly higher than their nonverbal score. Only 6 participants (9%) had the opposite pattern, with significantly higher nonverbal abilities. Among school-aged children (> 6 years), nonverbal skills could be further broken down into two clusters: spatial skills and nonverbal reasoning. Differences between spatial and nonverbal reasoning skills did not reach levels of significance, t(50) = 1.1, p = .28.
Several analyses were conducted in order to explore whether genetic differences could account for variability in cognitive functioning in NS. The cohort was first examined based on the gene in which a mutation was found. The single participant with a BRAF mutation was not included in these analyses. Figure 2 depicts the distribution of full-scale DAS scores for individuals with PTPN11 mutations, SOS1 mutations and unknown mutations. The “unknown” mutation group was expected to be heterogeneous with respect to the disease-causing gene.
All six participants with SOS1 mutations scored within the average range or higher on the cognitive assessment (range: 91–123). The average full-scale DAS performance for the SOS1 group did not differ from the normative population (100), t(5) = .74, p > .40. In contrast, more than half of individuals with PTPN11 mutations (n = 20; 61%) scored below the average range (> 90). Scores in this group ranged from 59–110. The mean DAS performance for the PTPN11 group was significantly lower than expected based on the normative population mean, t(32) = −6.3, p < .001. Similar to the PTPN11 group, the group of participants with unknown mutation status scored lower on the DAS than expected based on the population mean, t(24) = −3.59, p < .001. Scores for the unknown mutation group ranged from 44–123. The distribution of scores for the unknown group was less similar to a normal curve than the distributions for the PTPN11 genotype. Descriptive statistics (e.g., range and standard deviation) for this group were also larger, suggesting that this group had greater variability with respect to cognitive skills. Note that it is expected that approximately fifty percent of the 12 untested individuals in the unknown mutation group have a PTPN11 mutation and ten percent an SOS1 mutation.
Comparisons were conducted to determine whether cognitive functioning in NS patients differed based on the presence or absence of specific mutations. Because twelve participants in the unknown mutation group had not been tested for any NS genes, these individuals were excluded from the following analyses. Three groups remained: a PTPN11-positive group, an SOS1-positive group, and a group of PTPN11-negative individuals whose genotype is unknown (two of whom were also known to be SOS1-negative). A one-way ANOVA was conducted to examine whether there were reliable differences in intellectual ability among the three genotype groups. This analysis indicated a significant difference in full scale DAS scores between the groups, F(2,49) = 3.50, p < .05.
Several planned comparisons were conducted to examine these genotype differences in cognitive ability more closely. The first analysis compared the performance of the SOS1 group and the PTPN11 group. The SOS1 group scored significantly higher on the DAS than the PTPN11 group, t(37) = 3.16, p < .01. In order to examine whether the observed genotype difference was consistent across different domains of intellectual functioning, the PTPN11 and SOS1 groups were compared separately on the verbal cluster and the nonverbal cluster. Results were identical to the full-scale test. The SOS1 group performed significantly better than the PTPN11 group on both verbal, t(37) = 2.80, p < .01, and nonverbal, t(37) = 2.90, p < .01 cognitive scales (See Figure 3).
The SOS1 and PTPN11 mutation groups were then compared to individuals without mutations in each of those genes, respectively. An analysis was performed to contrast individuals with SOS1 mutations to an SOS1 mutation-negative group. The latter group included both PTPN11-positive individuals (with the assumption that these patients would not also be SOS1-positive) as well as individuals with unknown mutations who tested negative for SOS1 mutations. SOS1-positive individuals scored higher than SOS1-negative individuals on the DAS, t(39) = 2.73, p < .01. The PTPN11-positive group was compared with a PTPN11-negative group (including both the PTPN11-negative individuals of unknown genotype and the SOS1-positive participants). PTPN11-positive individuals scored significantly lower than the PTPN11-negative individuals on the DAS, t(50) = −2.09, p < .05. When the 6 participants with identified SOS1 mutations were removed from this analysis, however, the difference between the PTPN11-positive and PTPN11-negative groups was no longer significant, t(44) = −1.08, p = .29.
Additional analyses examined the PTPN11-positive group in greater depth. Table 2 displays DAS scores grouped by exon in which a mutation was detected. Group sizes were not sufficient to detect differences among these groups. However, a wide range of abilities was seen for each exon group; at least one participant in each group had low-borderline functioning, and at least one participant in each group scored in the average range or higher for a given cluster. Hence it appears that PTPN11 mutations across the whole gene have the potential to interfere with cognitive development. However, PTPN11 mutations in all exons are compatible with normal cognitive development.
For individuals with mutations in exon 8, the group mean was in the average range for both verbal and nonverbal skills. In a previous study (Tartaglia et al., 2002), 17 patients with an N308D mutation in exon 8 were all found to attend regular education classrooms. The four individuals with this mutation in our sample fit this profile of having no (or mild) cognitive delays. Three had full-scale DAS scores in the average range and one scored in the low average range (M = 92.5, SD = 6.3). In order to determine whether this finding could extend to all individuals with N308 mutations, we also examined the scores of individuals in our cohort with N308S and N308T mutations. The two patients with N308S mutations both received a full-scale score in the average range (SS = 104 & 101). The single individual with an N308T mutation scored in the range of mild mental retardation (SS = 61). Hence even within N308 mutations, variability was evident for different amino acid substitutions. Nevertheless, the absence of marked cognitive deficits among any individuals with N308D and N308S mutations in our sample suggests that some N308 mutations are likely to be associated with mild cognitive effects.
The individual with a BRAF mutation in our sample (age 14 years, 4 months) achieved an overall DAS score in the Low range (SS = 76; 5th percentile), with a verbal score in the Low Average range (SS = 85; 15th percentile) and a nonverbal score in the Low range (SS = 74; 4th percentile). Although scores in all domains were below the average range, mental retardation was not present. This participant scored higher than 24% of the individuals with PTPN11 mutations and 32% of individuals with unknown mutations on the DAS. Thus, observed scores for this BRAF-positive individual were within the range of scores seen in other NS patients.
In order to further probe potential influences on cognitive abilities in NS, four additional factors were also explored. These factors included two medical features associated with Noonan syndrome (severity of cardiac disease and hearing screening scores), a developmental factor (motor coordination) and a measure of socioeconomic status (years of parental education). Descriptive statistics for these factors are reported in Table 3. Multiple regression analyses were conducted to investigate whether these variables had significant influence on verbal skills and nonverbal skills. Three variables accounted for 34% of the variability in verbal intellectual functioning: hearing screening scores, manual motor dexterity and years of parental education (Table 4). Two variables, motor dexterity and parental education, were significantly predictive of nonverbal skills, accounting for 40% of the variability in DAS nonverbal scores. Severity of heart disease was not predictive of any of the cognitive outcomes.
In order to examine whether the observed genotype difference between the PTPN11 and SOS1 groups could be explained by group differences on these additional (non-genotypic) factors, we compared the scores of these two genotype groups on each factor. Mean scores for individuals with SOS1 and PTPN11 mutations on each factor are included in Table 3. Scores were not significantly different between the two groups for severity of cardiac disease, t(37) = 1.15, p = .26, hearing screening scores, t(37) = .11, p = .91, motor dexterity, t(28) = .02, p = .98, or parental education levels, t(37) = .43, p = .67. Thus the SOS1 group and PTPN11 group did not differ reliably on scores for any of the additional predictors of cognitive functioning that were measured.
Additional analyses were conducted to determine whether individuals in our sample who also had a parent with NS (who may have had learning disabilities that impacted educational attainment) were driving the association between parental education and cognitive ability. The parental education levels of participants with an affected parent and those without an affected parent were compared. Parent education levels of the 17 participants with likely inherited NS mutations (M = 15.56 years, SD = 1.2) did not differ significantly from the 48 patients with sporadic mutations (M = 15.63 years, SD = 2.1), t(62) = .13, p = .90. Individuals with familial NS also did not differ in levels of cognitive ability from those with sporadic mutations, t(63) = −.55, p = .58. In order to control for mutation type, we also examined differences between familial vs. sporadic cases among only those individuals with PTPN11 mutations. Differences in parental education level, t(31) = −.92, p = .36, and cognitive ability, t(31) = .67, p = .51, did not reach levels of significance. Hence parental education levels have a significant impact on their child’s cognitive ability independent of whether a parent also has a mutation; in addition, having a familial form of NS does not appear to pose additional cognitive risk.
Although multiple studies have established that cognitive impairments are more common in Noonan syndrome than in the general population, little is known about the causal pathways that lead to these outcomes. The current study is the first to explore whether genetic differences can explain some of the wide variation in cognitive functioning in individuals with NS.
As a group, the pattern of performance on cognitive assessments among NS individuals in this study was similar to previous reports (Lee et al., 2005; Van Der Burgt et al., 1999). Cognitive disabilities were present with greater frequency in NS than in the general population, and the overall distribution of cognitive test scores was shifted significantly downward. Nevertheless, a large proportion of individuals in this sample (~50%) demonstrated intellectual skills in the average range or higher.
Cognitive scores in our cohort did not vary significantly as a function of the chronological age or gender of participants, or based on whether the mutation was sporadic vs. familial. On average, individuals in our sample had significantly better verbal abilities than nonverbal abilities, a finding that is consistent with one study (Van Der Burgt et al., 1999), but opposite of the pattern found in another (Lee et al., 2005). This inconsistency in patterns could reflect differences in the measurement tools used. For example, the school-age nonverbal scale in the DAS contains a test that requires participants to draw complex spatial figures from memory (Recall of Designs). Individuals with NS in our study achieved lower scores on this subtest than any other, perhaps because this subtest relies somewhat on fine motor skills. Manual fine motor skills were severely impaired (> 2 SD below the mean) in 34% of those tested on the Purdue Pegboard Task and below average (> 1 SD below the mean) in 72% of the sample. Hence the greater reliance on motor skills for nonverbal tests could potentially account for the verbal advantage in our sample. This explanation does not account for the discrepant results in the previous studies, which were conducted using two versions of the same (Wechsler Intelligence) scale, the WISC-R and the WISC-RN. An alternative explanation for the inconsistency in patterns of verbal-nonverbal ability is that the sample sizes available for each study were relatively small, which may lead to some instability in outcomes. Nevertheless, taken together, studies of cognitive functioning in NS suggest that a consistent pattern of cognitive strengths and weaknesses is unlikely to emerge in NS.
Our genotype-phenotype analyses indicate that some of the variation in cognitive skills in NS is attributable to differences in genotype. Individuals with SOS1 mutations performed significantly higher on both verbal and nonverbal cognitive tests than individuals with PTPN11 mutations and SOS1-negative individuals with unknown mutations. Although limited research has been conducted to examine individuals with SOS1 mutations, this finding is consistent with reports that these individuals are more likely to be placed in regular education classrooms than those with PTPN11 mutations (Tartaglia et al., 2007). Our results also support previous reports (Tartaglia et al., 2002) that individuals with the N308D mutation in PTPN11 are likely to have no or mild cognitive disabilities.
The cognitive differences between individuals with SOS1 and PTPN11 mutations in this sample were seen in both verbal and nonverbal domains, suggesting that the group difference is not due to a specific domain of strength or weakness caused by mutations in a particular gene. Age differences also cannot account for the discrepancy, as the two groups did not differ significantly in chronological age, t(37) = 1.15, p = .26, and age was uncorrelated with performance on the cognitive tests. Note that this sample did not include infants and toddlers under age 4, for whom developmental delays have been reported in SOS1-positive (Narumi et al., 2008) as well as PTPN11-positive individuals. The difference between the groups was also not due to differences in rates of heart disease, hearing loss, motor incoordination or socioeconomic level, suggesting that some other factor is responsible for the differences in cognitive performance.
One possible explanation for the group differences in cognitive performance is that alterations in SOS1 gene expression have less impact on central nervous system development than alterations in other RAS-MAP Kinase genes. Another possibility is that other unidentified medical or developmental factors vary across the genotypes, and one or more of these factors are affecting cognitive outcomes. A third possibility is that some individuals with SOS1 mutations do have cognitive disabilities, but these individuals may carry a diagnosis other than NS and therefore would not have been recruited for the current study. It has been recognized that many SOS1 patients have unusual ectodermal features similar to those seen in CFC syndrome (Tartaglia et al., 2007). Given the substantial overlap in physical features of NS patients with SOS1 mutations and CFC syndrome patients (Narumi et al., 2008), this possibility should be further explored. Indeed, replication of our genotype-phenotype results with a larger sample of SOS1 individuals is necessary for distinguishing among these possibilities.
Our sample included only one individual with a BRAF mutation, presumably because people with this genotype typically carry a CFC syndrome diagnosis rather than NS (Rodriguez-Viciana et al., 2006). This case is notable due to the fact that the individual we assessed did not have mental retardation, although mental retardation has been reported to be universally present in CFC syndrome (Armour & Allanson, 2008; Yoon et al., 2007). Our BRAF-positive participant also lacked the ectodermal features that are present among most individuals with a CFC diagnosis. She did display below average cognitive skills; however, her verbal ability was near the average range. In addition, this person achieved scores well within the range seen in other individuals with Noonan syndrome in our sample. This individual provides further evidence for an overlap in the NS and CFC phenotypes, even among individuals with known mutations (see also Nystrom et al., 2008). In addition, although mental retardation is common among BRAF-positive individuals, our study suggests that mental retardation is not a necessary consequence of all BRAF mutations. Further research is needed to delineate whether differences in cognitive phenotype can be linked to specific mutations in BRAF.
In addition to the genetic differences observed in this study, we also investigated whether additional medical, developmental or environmental variables accounted for variation in cognitive ability. Severity of cardiac disease was not associated with cognitive functioning in this population. However, failure to pass a hearing screening was significantly associated with lower performance on verbal tests. In addition, verbal and nonverbal cognitive abilities were significantly predicted by the socioeconomic status and fine motor abilities of the children in our study. These analyses indicate that not only genetic factors, but also developmental and social factors play a significant role in cognitive development in NS.
Studies of genotype-phenotype correlations in the neuropsychological realm, which have been enabled by important advances in molecular genetics, have significant implications for clinical care. If the genotype is known for a given patient, specific avenues for early intervention and other educational planning may be indicated. Our research suggests that individuals with PTPN11 mutations and unknown mutations are at risk for cognitive disabilities, although a wide range of abilities was observed. Comprehensive neuropsychological testing can help to identify areas that require special attention. NS patients with SOS1 mutations and N308D/N308S mutations in PTPN11 appear likely to develop normal range cognitive skills, although functioning in other areas (e.g., motor development) may be delayed.
It is important to note that the substantial overlap in the distributions of cognitive scores between all NS genotypes suggests that having a specific genetic anomaly does not indicate reliably what the cognitive outcome will be for any given individual. Indeed, the significant influence of several non-genotypic factors suggests that a number of steps can be taken to foster development in NS regardless of genotype. For all patients, physical or occupational therapies for motor impairments may improve achievement on cognitive as well as physical tests. Adaptations for individuals with motor difficulties can be implemented in educational settings so that these disabilities do not affect achievement in other areas. In addition, identification of hearing impairments and subsequent intervention are of great importance to enhance the development of verbal skills. It is critical that clinicians and school professionals are made aware of the range of issues associated with NS so that all proper evaluations and modifications can be administered.
This study represents a first step toward examining the differential effects of RAS/MAP Kinase pathway gene mutations on measurable cognitive behaviors. Further research is needed to examine other aspects of learning and behavior in Noonan syndrome, and to determine how these characteristics relate to aspects of the medical and genetic history of affected individuals. Establishing genotype-phenotype relations in the neuropsychological realm may help to improve our knowledge of the impact of specific genes on the developing nervous system. However, this line of research has only begun to get underway. Identification of new NS genes and their roles in the RAS/MAP Kinase pathway is occurring at a rapid pace, and these advances will continue to add crucial pieces to the puzzle.
The authors are grateful to all of the participating families who made this research possible. Special thanks go to Wanda Robinson and The Noonan Syndrome Support Group (TNSSG) for their support. We would also like to thank Dr. Susan Ellis Weismer and Dr. Robert Nellis for their invaluable advice and suggestions, and Dr. Richard Pauli, Dr. David Wargowski and Jody Haun at the Waisman Center for their assistance in recruiting participants. The work was supported by NIH grant T32 HD049899-01 and a Royalty Foundation Research Fellowship.