Neurocognitive Dysfunction in

Children with Neurofibromatosis Type 1

Tena L. Rosser, MD, and Roger J. Packer, MD

Address
Department of Neurology, Children's National Medical Center,
111 Michigan Avenue NW, Washington, DC 20010, USA.
E-mail: trosser@cnmc.org
Current Neurology and Neuroscience Reports 2003, 3:129136
Current Science Inc. ISSN 15284042
Copyright 2003 by Current Science Inc.

The cognitive dysfunction associated with neurofibromatosis type 1 (NF1) is an intriguing aspect of this phenotypically heterogeneous genetic neurocutaneous disorder. A
broad range of both nonverbal and verbal learning disabilities are evident in approximately 30% to 65% of children
with NF1. Deficits in IQ, executive function, attention,
and motor skills have also been documented. Current
challenges lie in discovering the underlying multifactorial
etiologies of the cognitive abnormalities found in NF1.
Likely answers lie in neuroanatomic correlates as seen on
neuroimaging as well as in molecular and genetic advances
into the role of neurofibromin, the protein product of the
NF1 gene. The development of NF1 animal models with
learning and memory difficulties similar to those seen in
humans demonstrates promising preliminary evidence
that medical treatment of cognitive abnormalities may one
day be possible.

Introduction
Neurofibromatosis type 1 (NF1) is a common multisystem disorder of autosomal dominant inheritance that
affects approximately one in 3000 individuals of all ethnic groups [1]. The physical manifestations of NF1, such
as caf au lait spots, axillary freckling, iris hamartomas
(Lisch nodules), osseous lesions (sphenoid wing dysplasia, pseudarthrosis) and benign as well as malignant neural tumors (neurofibromas, optic gliomas), are well
recognized (Table 1) [2]. In addition, cognitive difficulties and learning disabilities of varying degrees have been
noted in 30% to 65% of patients [3]. The cognitive dysfunction associated with NF1 is a complicated and
incompletely understood aspect of the disease. It also
represents an important source of morbidity, as an individual's cognitive abilities have a major lifelong impact
on educational prospects, future employment, selfimage, relationships with peers, and overall ability to
function in society.

Over the past decade, much research has been devoted

to unraveling the mystery of cognitive dysfunction in
NF1, with the goals of defining the specific deficits, determining the underlying causes, and eventually discovering
effective treatments. Modernization of neuroimaging
methods and genetic techniques has made significant
advances in this realm possible. Important recent developments have included expanded research into the relationship between brain structure and cognitive function,
better understanding of the NF1 gene on chromosome
17, identification of several functions of neurofibromin
(the protein product of the NF1 gene), and the development of animal models with cognitive deficits similar to
those in humans with NF1.

Background: What Do We Already Know?

Mental retardation and IQ
Recent data based on objective testing measures have
revealed that the NF1 population has a slightly increased
incidence of mental retardation, defined as a full-scale IQ
score of less than 70. Between 4% and 8% of individuals
with NF1 fall into the mentally retarded IQ range, whereas
approximately 3% of the general population has mental
retardation [3,4]. Across numerous studies of the NF1 population, mean full-scale IQ scores vary from 87 to 100 on
Wechsler Intelligence Scales (WISC) [5]. The vast majority
of studies have shown no discrepancy between performance and verbal IQ scores [613].
Debate remains regarding whether the lowering of IQ
applies to the NF1 population as a whole or if it represents only a subset of NF1 patients [4,5]. There is some
preliminary evidence to suggest that a bimodal pattern
exists, indicating that there are two distinct groups, those
with cognitive impairment and those without [6,14,15].
However, further research is required to substantiate
these suspicions.
Visual-spatial skills
Deficits in visual-spatial skills are well documented in NF1.
Since the 1980s, when these difficulties were first identified, multiple studies have evaluated visual-spatial skills
using various measures including Bender visual-motor
gestalt testing, dot localization, facial recognition, Hooper
visual organizational testing, judgement of line orientation
(JLO), Rey-Osterreith complex figure copying, visual-

130

Pediatric Neurology

motor integration, visual form discrimination, visual-perceptual skills testing, as well as others [6,7,1012,1618].
The JLO test stands out among studies as being consistently abnormal [4,6,7,9,10,1618]. Thus, the JLO is
believed to be an important indicator of neurocognitive
dysfunction in NF1 [4].
Two studies have documented that the visual-spatial
deficits found in NF1 are somewhat unique. Cutting et
al. [19], expanding initial research by Eliason [16],
compared children with NF1 with learning-disabled
children without NF1 and with a normal control group.
The NF1 group scored significantly lower on visual-spatial tests than the learning-disabled group; however,
there was no significant difference between the NF1 and
the control groups. The authors concluded that the
visual-spatial deficits in children with NF1 are different
than those found in otherwise healthy learning-disabled children [19].

Learning disabilities and the cognitive phenotype

Reports of the frequency of learning disabilities (LD) in
NF1 range from 30% to 65%, which is significantly higher
than the estimates of 7% to 10% of LD seen in the general
population [3,19]. A learning disability represents a discrepancy between ability (IQ) and achievement (performance on reading, writing, spelling, or math testing) [4].
Both Ozonoff [5] and North [4] have noted in recent
review articles that estimates of learning disabilities in the
NF1 population require cautious interpretation secondary
to confounding factors. There are differing definitions of
LD, lack of appropriate control groups, and frequently no
accommodation for the lowering of IQ seen in the NF1
population among studies [4,5].
Cutting et al. [19] have made another interesting
point regarding the interpreting of studies of cognition.
When comparing three groups of children (NF1, learning disabled without NF1, and normal control groups),
several visual-spatial and language tasks previously
found to be deficient in sibling-matched designs were
not present when comparing the NF1 group and the nodiagnosis control group [19]. Thus, these authors have
emphasized that broad genetic and familial influences
must be taken into account when considering the influence of NF1 on cognition.
Early work into the learning disabilities associated with
NF1 attempted to describe an NF1 cognitive phenotype
[7]. Researchers now know that a spectrum of nonverbal as
well as verbal learning disabilities exist in children with
NF1. Thus, there does not appear to be a specific profile of
cognitive dysfunction in the NF1 population [3]. This realization has implications for treatment, as neurocognitive
evaluation and management must then be individualized
for every child with NF1 and LD.

Language function
Although the initial focus was on describing the nonverbal
deficits in NF1, it was presumed that language function
was relatively spared in these children. However, the vast
body of research performed most recently has demonstrated a variety of language impairments in the realms of
reading, spelling, vocabulary, naming, verbal reasoning,
wri t ten mathemat i cs, and wri tt en l anguage
[6,7,10,12,16,17,19]. Dyslexia, a specific reading disability,
has also been identified in children with NF1 in some
studies [7,10,17].
Verbal disabilities do not frequently occur alone in
children with NF1. They may be less severe than, but
coexist with, visual-spatial difficulties [4,5]. Eliason [16],
for example, evaluated 23 children with NF1. He identified isolated visual-perceptual disabilities in 56% and
concomitant language dysfunction in 30%. The exact
nature of the overlap between visual-spatial abnormalities and language dysfunction in children with NF1 is still
incompletely understood.

Executive functioning
Executive functioning requires goal-directed behavior and
the use of numerous complex intellectual skills, such as
organization, planning, attention, inhibition, and selfmonitoring [5].
Although less well studied than other aspects of the
cognitive dysfunction in NF1, executive functioning is
impaired in some individuals with NF1. The majority of
studies have focused on children with NF1. For example, in
the initial works of the 1980s, Stine and Adams [20] noted
"weaknesses related to cortical organization and distractibility." [20]. Later, others revealed poor problem solving
skills, strategy generation, and expression of ideas [14].
There is also evidence to suggest that these difficulties
persist into adulthood. In evaluating adults with NF1, Zoller et al. [21] found that NF1 had a negative impact on a
variety of intellectual skills, such as inductive reasoning,
logical abstraction, short-term memory, attention, and
mental flexibility. This raises the question of whether specific interventions during childhood could impact these

Table 1. National Institutes of Health diagnostic

criteria for NF1
Two or more of the following features must be present for
the diagnosis of NF1:
6 or more caf au lait spots with a diameter greater than 0.5
cm before puberty and 1.5 cm after puberty
Axillary or inguinal freckling
2 or more neurofibromas or a single plexiform
neurofibroma
2 or more Lisch nodules (iris hamartomas)
Optic glioma
A distinctive osseous lesion such as sphenoid bone dysplasia
or thinning of long bone cortex
A first-degree relative with NF1 by the above criteria
NF1neurofibromatosis type 1.

Neurocognitive Dysfunction in Children with Neurofibromatosis Type 1 Rosser and Packer

131

Table 2. Results of studies on ADHD in children with NF1

Study/year

Patients with NF1 children

NF1, n
with ADHD* Comparison population Conclusions

Kayl et al. [23] /

2000

36

33%

Koth et al. [25] /

2000

31

42%

Mautner et al.
[24] / 2002

93

49.50%

18 normal control subjects NF1 patients had increased total corpus

callosum area compared with control
subjects, but there were no differences
between NF1 and NF1+ADHD groups.
Increased severity of attention difficulties
in NF1+ADHD group was related to
smaller splenial and total callosal area.
NF1-unaffected siblings and ADHD may be a component of NF1
biologic parents
phenotype rather than a separate,
unrelated disorder.
26 children with NF1 + no First systematic evaluation of effectiveness
of stimulant medication in 20 children
ADHD, 14 children with
with NF1+ADHD. Showed
ADHD + no NF1, 14
methylphenidate improved attention and
control subjects
impulse control on TOVA test.
Methylphenidate also improved learning
ability and social behavior at 1 year of
follow-up by the CBCL.

*Children who meet diagnostic criteria for ADHD by DSM III-R of DSM-IV criteria.
ADHDattention deficit hyperactivity disorder; CBCLchild behavior checklist; NF1neurofibromatosis type 1;
TOVAtest of variables of attention.

mental functions later in life and underscores the importance of further research into this area of cognition.
Attention deficits
Attention has been one of the better-analyzed aspects of
executive functioning in the NF1 population and is an area
of keen interest in NF1 research at this time (Table 2). As
noted previously, multiple studies have suggested that
there is an association between NF1 and attention deficit
hyperactivity disorder (ADHD), defined by the DSM-IV criteria as a constellation of inattention, hyperactivity, and
impulsivity [13,14,16,2022]. Some researchers have
shown, however, that children with NF1 may be more
likely to fulfill diagnostic criteria for attention deficit disorder without hyperactivity [14].
Although the exact incidence is unknown, recent work
suggests that ADHD may be present in approximately one
third to one half of children with NF1, a marked increase
above the estimates of ADHD in the general pediatric population (5%) [7,23,24]. More specifically, in agreement
with earlier reports by Hofman et al. [7], Kayl et al. [23]
found that 12 of 36 (33%) of their NF1 pediatric study
population met diagnostic criteria for ADHD by rating
scales. Mautner et al. [24] have found an even higher
incidence, documenting that 46 of 93 (49.5%) of their
pediatric NF1 study population satisfied DSM-IV ADHD
diagnostic criteria.
Koth et al. [25] have raised the question of whether
ADHD is part of the NF1 phenotype or a separate, unrelated disorder. Controlling for genetic and environmental
factors by comparing the ADHD status of children with
NF1 to their unaffected siblings and biologic parents, the

authors findings support the theory that ADHD is indeed a

component of the NF1 behavioral and cognitive phenotype [25].
Motor skills
Both gross and fine motor delays are commonly found in
children with NF1. A determination of the precise incidence of motor delays in the pediatric NF1 population is
difficult to discern from the literature. However, deficits of
manual dexterity, coordination, balance, gait, and ball
skills have been noted in children with NF1 when compared with published age-related norms, unaffected siblings, and IQ-matched learning-disabled control subjects
[7,14,17]. Psychomotor slowing on motor tasks has also
been demonstrated [26]. These motor impairments may
impact upon a child's ability to interact with their environment and, as with cognitive delays, warrant prompt evaluation and intervention. Table 3 summarizes the
neurocognitive deficits in children with NF1.

Relating Structure to Function

Although many researchers have focused on defining the
neurocognitive deficiencies in NF1, others have investigated the possible underlying mechanisms by attempting
to relate brain function to its structure (Table 4). Subtle
abnormalities of brain structure in NF1, such as macrocephaly, which occurs in approximately 50% of individuals
with NF1, were recognized early in the study of this disease
[1]. The advent of modern neuroimaging with magnetic
resonance imaging (MRI) led to the discovery of T2weighted signal abnormalities or unidentified bright

132

Pediatric Neurology

Table 3. Summary of neurocognitive deficits in children with NF1

Realms affected

Comments

Mental retardation

Full-scale IQ less than 70 occurs in 4%8% of individuals with NF1. No consistent discrepancy between
performance and verbal IQ among studies. IQ may follow a bimodal pattern in NF1.
The most common type of neurocognitive deficit. JLO test is consistently abnormal between studies.
May be different in children with NF1 and unaffected children with LD.
A variety of verbal impairments may be seen. Usually coexist with visual-spatial difficulties.
Occur in 30%65% of children with NF1, a marked increase from the incidence of LD in the
general population.
Children with NF1 have a range of nonverbal and verbal disabilities. At this time, a specific NF1
cognitive phenotype is not recognized.
Deficits of goal-directed behavior and intellectual skills such as organization, planning, attention,
inhibition, and self-monitoring have been noted but are less well characterized than other aspects of
cognitive dysfunction in NF1.
Recent research shows that ADHD or ADD may occur in 33%49.5% of children with NF1. Stimulant
medications are effective in controlling symptoms.
Gross and fine motor delays as well as psychomotor slowing are common.

Visual-spatial skills
Language skills
Learning disabilities
Cognitive phenotype
Executive functioning
Attention
Motor skills

ADDattention deficit disorder; ADHDattention deficit hyperactivity disorder; JLOjudgement of line orientation;
LDlearning disability; NF1neurofibromatosis type 1.

objects (UBOs) and to the development of morphometric

and volumetric brain analyses, which have been helpful in
further advancing our understanding of cognition in NF1.
Functional MRI will likely bring the next wave of new
insights in this area.
Unidentified bright objects and magnetic resonance
imaging in neurofibromatosis type 1
Unidentified bright objects represent areas of high T2weighted signal on MRI and, by definition, demonstrate
no mass effect, edema, or enhancement with contrast (Fig.
1) [1]. Only one case report has linked a UBO to a focal
neurologic abnormality. A 13-year-old girl with NF1 and a
thalamic/subthalamic T2 signal hyperintensity was
reported to have a contralateral hand dystonia that
improved over 2 years as her radiologic abnormality
resolved [27]. UBOs occur primarily in the basal ganglia,
cerebellum, brainstem, and subcortical white matter of
approximately 43% to 79% of children with NF1 [4]. In
addition, UBOs are not static in nature, typically resolving
by early adulthood [3]. This may partially explain the wide
range of reported incidence of UBOs in the literature [3].
Histologically, UBOs demonstrate dysplastic or hyperplastic glial proliferation and white matter spongiform change
that is believed to represent intramyelinic edema. It has
been postulated that myelin abnormalities may then be at
least partially responsible for the cognitive difficulties seen
in NF1 [28].
A tremendous amount of research has been devoted
to determining how cognition might be related to the
number, size, volume, and precise location of UBOs,
with little consensus reached among studies. Initial
work by Dunn and Roos [29], Duffner et al. [30], and
Ferner et al. [31] showed no association between T2
hyperintensities and cognition, but this work was critically flawed by not performing IQ testing in all patients.

However, better designed follow-up studies using more

systematic, quantitative neurocognitive evaluations also
produced conflicting data. The results of several of these
studies are discussed to demonstrate the difficulty in
drawing firm conclusions from over a decade of research
in this area.
North et al. [6] showed lowering of IQ as well as
language and visual-spatial deficits in children with T2
hyperintensities. Lowering of global IQ in addition to
deficiencies of attention, visual-spatial skills, and executive functioning were demonstrated in children with
UBOs in the work of Joy et al. [9]. The number of UBOs
correlated with IQ deficits in a study by Hofman et al.
[7]. Denckla et al. [32] found that lower IQ scores in
children with NF1 were related to the presence and distribution of UBOs [32]. In contrast, work performed by
Legius et al. [33] and Moore et al. [13] showed no significant difference in IQ scores of the T2-positive and
T2-negative groups, although lower IQ was associated
with T2 lesions of the thalamus in the study by Moore
et al. [13].
In summary, evidence suggests that a relationship
between T2 hyperintensities and cognition in NF1 exists,
but the precise impact of UBOs on different areas of cognition is still controversial. MRI T2 hyperintensities, thus,
cannot be considered a radiographic marker or predictor of
cognitive dysfunction in NF1 at this time [4].
Other neuroanatomic correlates
Questions regarding how other structural aspects of the
brain in individuals with NF1 may be linked to cognitive function have also arisen. Morphometric and volumetric MRI have made exploration of the phenomenon
of macrocephaly, as well as assessment of corpus callosal area and sylvian fissure morphology in the NF1
population, possible.

Neurocognitive Dysfunction in Children with Neurofibromatosis Type 1 Rosser and Packer

133

Table 4. Neuroanatomic structures possibly related

to cognitive dysfunction in NF1
Presence, location, and number of unidentified
bright objects
Macrocephaly/megalencephaly
Corpus callosal area
Planum temporale
NF1neurofibromatosis type 1.

Preliminary work has demonstrated that brain volume is significantly larger in children with NF1
compared with control subjects, suggesting that megalencephaly is responsible for macrocephaly in these patients.
There is conflicting evidence, however, as to how macrocephaly correlates with cognitive dysfunction [18,34,35].
Although Ferner et al. [26] and North et al. [6] found no
association, Cutting et al. [35] found significant vocabulary impairments in a macrocephalic NF1 population
compared with a normocephalic NF1 population. Of
note though, Cutting et al. [35] did not identify a strong
relationship between macrocephaly and the origin of NF1
(familial versus sporadic), UBOs (presence versus
absence), or extent of cognitive impairment.
Moore et al. [18] undertook further study of this subject, finding that gray matter volume was greater in children with NF1 compared with control subjects and
correlated with the degree of learning disability in the
NF1 study population. Corpus callosum size was also
larger in NF1 subjects and was related to lower scores on
measures of academic achievement in addition to visualspatial and motor skills [18]. The authors conclude by
proposing that delayed neuronal apoptosis may be
responsible for macrocephaly and delayed neuronal connections in children with NF1, suggesting that these
underlying structural abnormalities are related to cognitive impairment in NF1 [18].
Billingsley et al. [36] speculated that individuals
with NF1 might demonstrate cortical abnormalities similar to those seen in the reading-disabled general population, becoming the first to evaluate sylvian fissure
morphology specifically in children with NF1. Several
intriguing findings regarding planum temporale (PT)
structure were documented. The left PT in boys had a
smaller surface area and gray matter volume compared
with NF1 girls and control subjects. In addition, boys
with NF1 showed greater left-right PT symmetry, similar
to what is reported in the general reading-disabled population. Reading and math intelligence-based discrepancy
scores were related to PT asymmetry in the NF1 group
compared with control subjects. Less leftward PT asymmetry in the NF1 population was associated with poorer
reading and math achievement relative to full-scale IQ.
The authors conclude that the susceptibility of reading
and other learning disabilities in NF1 may be related to
sylvian fissure development and morphology [36].

Figure 1. Brain magnetic resonance imaging fluid attenuated inversion recovery image demonstrates unidentified bright objects in the
white matter of the cerebellum, middle cerebellar peduncle, and pons
of a 3-year-old girl with neurofibromatosis type 1.

Attention deficit hyperactivity disorder and brain

morphology in neurofibromatosis type 1
Research in non-NF1 populations suggests that the underlying etiology of ADHD can at least partially be explained
by brain morphology [23]. Several recent studies have specifically examined the correlation between brain morphology and ADHD in children with NF1.
A recent MRI study by Cutting et al. [37] suggests that
megalencephaly in NF1 may be related to ADHD. Megalencephaly was present in boys with NF1 and no ADHD. In
contrast, boys with NF1 and ADHD were normocephalic,
suggesting that somehow a diagnosis of ADHD in NF1 corrects for megalencephaly [37].
A relationship between anterior and posterior corpus
callosum structure and ADHD versus normal control subjects is also believed to exist [23]. Kayl et al. [23] evaluated
the cross-sectional areas of seven regions of the corpus callosum in children with NF1 to determine how morphology was related to ADHD symptoms. The authors
demonstrated that children with NF1 had increased corpus
callosal areas over control subjects, but there were no discernable differences in callosal area between the NF1 and
NF1/ADHD groups. However, unexpectedly, an increase in
severity of attention deficits was associated with a smaller
splenial and total corpus callosum area in the NF1/ADHD
children [23].

Management of Learning-disabled Children

with Neurofibromatosis Type 1
Although no well-defined NF1 cognitive phenotype has
emerged, it is currently understood that individuals with
NF1 may experience a variety of nonverbal and verbal neurocognitive deficits that can adversely affect their educational experience and have a large impact on their overall

134

Pediatric Neurology

quality of life. Cognitive dysfunction in children may at

times be subtle and may often masquerade as behavioral
problems, attention difficulties, or laziness. Thus, physicians and other individuals involved in the care of children
with NF1 should have a very high index of suspicion for
learning disabilities. Formal neurocognitive evaluation is
warranted in any child who shows possible cognitive
impairment. Gross and fine motor delays deserve the same
attention, as intervention with physical and occupational
therapy can improve motor skills and coordination.
Recognition and diagnosis of comorbid ADHD in children with NF1 is also critically important because effective
treatment with stimulant medication and behavioral modification techniques are available. Mautner et al. [24]
have published the first systematic study of the use of stimulant medication in the NF1 pediatric population, substantiating suspicions of effectiveness that were previously
primarily anecdotal in nature. Low-dose methylphenidate
(range, 5 to 15 mg; average of 7.5 mg) improved attention,
behavior control, and social problems in the NF1/ADHD
study population. At 1-year follow-up, these benefits were
still demonstrable and statistically significant on both parent and teacher Child Behavior Checklists [24]. Thus,
used in conjunction with nonmedical therapies such as
behavioral modification, stimulant medications may help
maximize academic as well as social abilities in children
with concomitant NF1 and ADHD.
Anticipatory guidance with regard to physical as well as
cognitive complications should also be provided to parents
of children with NF1 so that prompt intervention may be
undertaken when problems arise [3]. Specific educational
protocols are not currently available for learning-disabled
children with NF1, but they may be developed in the
future as the cognitive dysfunction associated with NF1 is
better understood.

Pathophysiology and the

Role of Neurofibromin
Understanding the underlying genetic defect of NF1 has
become critical to explaining the spectrum of clinical manifestations of this disease, from benign and malignant
tumor growth to cognitive dysfunction. Neurofibromin is
the protein product of the NF1 gene, which is located on
chromosome 17q11.2 [38]. In adults, neurofibromin is
expressed in neurons, Schwann cells, and oligodendrocytes, as well as in many non-neural cell types [39].
Although the function of neurofibromin is incompletely
understood, it clearly plays a role in multiple intracellular
processes, including regulation of the ras tumor suppressor
gene, modulation of adenylyl cyclase and cyclic adenosine
monophosphate (cAMP) pathways, as well as cytoskeletal
assembly [40,4144].
The effect of neurofibromin on the ras gene has been
most clearly defined. More specifically, a domain of the
NF1 gene contains GTPase-activating protein (GAP) activ-

ity which down-regulates ras, a plasma membrane protein

involved in cellular growth and differentiation [41]. The ras
protein has active and inactive forms. Ras is active when
bound to GTP and inactive when GTP is hydrolyzed to
GDP [45]. In the active state, ras promotes cell proliferation. GAP proteins, such as neurofibromin, regulate ras by
stimulating the conversion of active to inactive ras. Thus,
mutations in the NF1 gene produce inactive neurofibromin, and the resulting deficiency of normal intracellular
neurofibromin causes excessive ras activity with subsequent unregulated cell growth [46].

Animal Models of Cognitive

Dysfunction in Neurofibromatosis Type 1
Mouse models
The development of knockout mouse models has allowed
researchers to apply their knowledge of the pathogenesis of
NF1 and of the functions of neurofibromin to investigate
cognition dysfunction at molecular and cellular levels.
Building on the initial work of Silva et al. [47], two recent
articles published by Costa et al. [48,49] have been
instrumental in specifically linking the learning process in
NF1 to the interaction between neurofibromin and ras.
Initial mouse model studies demonstrated that mice
homozygous for a null mutation in the NF1 gene showed
mid-gestation lethality. However, mice heterozygous for
the same null mutation were viable, showing an increased
predisposition to tumor formation as well as learning deficits [47,48]. It was also determined that an alternatively
spliced exon 23a modifies the GAP domain of the NF1
gene [48]. Exclusion of exon 23a produces the type I isoform of neurofibromin, whereas inclusion of the exon 23a
results in the type II isoform of neurofibromin. To further
investigate the function of the type II isoform, Costa et al.
[48] developed a mouse model specifically lacking exon
23a (NF1 23a-/-). They found that although mice homozygous for this mutation survived without physical impairment or increased tumor predisposition, specific learning
deficits were apparent. Cognitive deficits were most notable on hippocampal-dependent tasks, such as the Morris
water maze and contextual discrimination, and were similar to those seen earlier in heterozygous null mice. The
authors concluded that the type II isoform of neurofibromin is essential for normal cognitive function and that the
GAP-related domain of neurofibromin plays an important
role in learning and memory. In addition, they proposed
that hyperactive ras may have been responsible for the NF1
exon 23a-/- learning-disabled phenotype [48].
Research has suggested that abnormal up-regulation of
ras activity, caused by loss of neurofibromin, may play a
crucial role in cognitive dysfunction in mice, as well as
humans, with NF1 [49]. In their second paper, to better
define the function of the ras protein in hippocampaldependent visual-spatial task learning, Costa et al. [49]
evaluated mice heterozygous for the NF1 gene null muta-

Neurocognitive Dysfunction in Children with Neurofibromatosis Type 1 Rosser and Packer

tion using a water maze test. The authors were able to demonstrate improved spatial learning deficits in NF1 mice by
manipulating ras activity genetically as well as pharmacologically. Specifically, NF1 mice (with presumed increased
ras activity) crossed with mice genetically engineered to be
deficient in ras activity did not have visual-spatial learning
difficulties. Also, administration of an oral farnesyl-transferase inhibitor, a medication that decreases ras signaling
by altering the post-translational step of farnesylation, to
adult NF1 mice corrected their learning deficits [49].
This study further evaluated the long-term potentiation
(LTP) of hippocampal CA1 neurons in these mice, and
found that the NF1 mice had a deficit of synaptic plasticity
[49]. LTP is a form of synaptic plasticity that is believed
to be involved in hippocampal-related learning paradigms
[40]. As with the learning deficits, the genetically engineered NF1 mice with deficient ras activity had normal
LTP, suggesting that hyperactive ras may also contribute to
decreased synaptic plasticity [40,49]. Furthermore, the
authors showed that increased gamma-aminobutyric acid
(GABA)-mediated inhibition may cause decreased synaptic
plasticity, and thus may be related to the underlying cognitive difficulties associated with NF1 [49].
Not yet testable in humans, these experiments have
outlined certain important parallels between the cognitive
deficits in mice and humans with NF1 and have relevant
implications for the development of treatment strategies
for cognitive dysfunction through modification of ras
activity or GABA mediation [49].

impairments, in children with NF1. Research into the multiple functions of neurofibromin and the development of
NF1 animal models with similar learning and memory disabilities have laid the foundation for explaining cognitive
dysfunction on molecular and cellular levels. Targeting
specific biochemical pathways to improve cognitive abilities both genetically and pharmacologically has now been
realized in mouse models and holds potential for similar
treatments in humans. Importantly, the developments in
our understanding are not just relevant to NF1, but have
implications for unlocking the mysteries of cognitive function in humankind.

References and Recommended Reading

Papers of particular interest, published recently, have been
highlighted as:

Of importance
Of major importance
1.
2.

3.

4.
5.
6.

Drosophila models
The NF1 protein product in Drosophila, the fruit fly, contains a 2802-amino acid sequence that is 60% identical to
that of human neurofibromin [42]. Thus, the Drosophila
model has been useful in exploring additional mechanisms involved in learning and memory in NF1.
Guo et al. [50] demonstrated that G-proteinactivated
adenylyl cyclase activity appears to have NF1-dependent
and NF1-independent components in Drosophila. The NF1dependent activation of the rut-adenylyl cyclase pathway
was required for olfactory learning and memory in the NF1
Drosophila, demonstrating a novel mechanism for how Gproteins activate cAMP pathways during the normal learning process [50]. Although it is not yet clear how these new
insights apply to humans, they again highlight the fact that
neurofibromin's role in cognitive function is extremely
complicated and that multiple biochemical pathways are
most likely involved.

Conclusions
Although many questions regarding cognition in children
with NF1 remain, we have begun to understand the complex mechanisms at play over the past two decades, making
great progress in defining a range of nonverbal and verbal
cognitive dysfunction, as well as behavioral and motor

135

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

North KN: Neurofibromatosis 1 in childhood. Sem Pediatr

Neurol 1998, 5:231242.
National Institutes of Health Consensus Development
Conference: Neurofibromatosis: Conference Statement
1988, 45:575578.
North KN, Riccardi V, Samango-Sprouse C, et al.: Cognitive
function and academic performance in neurofibromatosis 1:
Consensus statement from the NF1 cognitive disorders task
force. Neurology 1997, 48:11211127.
North K: Neurofibromatosis type 1. Am J Med Genet
2000, 97:119127.
Ozonoff S: Cognitive impairment in neurofibromatosis type
1. Am J Med Genet 1999, 89:4552.
North K, Joy P, Yuille D, et al.: Specific learning disability in
children with neurofibromatosis type 1: significance of MRI
abnormalities. Neurology 1994, 44:878883.
Hofman KJ, Harris EL, Bryan RN, Denckla MB:
Neurofibromatosis type 1: the cognitive phenotype.
J Pediatr 1994, 124:S1S8.
Moore BD, Ater JL, Needle MN, et al.: Neuropsychological
profile of children with neurofibromatosis, brain tumor, or
both. J Child Neurol 1994, 9:368377.
Joy P, Roberts C, North K, De Silva M: Neuropsychological
function and MRI abnormalities in neurofibromatosis type
1. Dev Med Child Neurol 1995, 37:906914.
Mazzocco MM, Turner JE, Denckla MB, et al.: Language and
reading deficits associated with neurofibromatosis type 1:
evidence for not-so-nonverbal learning disability. Dev Neurosci 1995, 11:503522.
Bawden H, Dooley J, Buckley D, et al.: MRI and nonverbal
cognitive deficits in children with neurofibromatosis 1. J Clin
Exp Neuropsychol 1996, 18:784792.
Dilts CV, Carey JC, Kircher JC, et al.: Children and adolescents
with neurofibromatosis 1: a behavioral phenotype. Dev Behav
Pediatr 1996, 17:229239.
Moore BD, Slopis JM, Schomer D, Jackson EF, Levy BM: Neuropsychological significance of areas of high signal intensity
on brain MRIs of children with neurofibromatosis. Neurology
1996, 46:16601668.
North K, Joy P, Yuille D, et al.: Cognitive function and
academic performance in children with neurofibromatosis
type 1. Dev Med Child Neurol 1995, 37:427436.
Brewer VR, Moore BD, Hiscock M: Learning disability subtypes in children with neurofibromatosis. J Learn Disabil
1997, 30:521533.
Eliason MJ: Neurofibromatosis: implications for learning and
behavior. Dev Behav Pediatr 1986, 7:175179.

136

17.

Pediatric Neurology

Eldridge R, Denckla MB, Bien E, et al.: Neurofibromatosis type

1 (Recklinghausens disease): neurologic and cognitive
assessment with sibling controls. Am J Dis Child
1989, 143:833837.
18. Moore BD, Slopis JM, Jackson EF, et al.: Brain volume in
children with neurofibromatosis type 1: relation to neuropsychological status. Neurology 2000, 54:914920.
19. Cutting LE, Koth CW, Denckla MB: How children with neurofibromatosis type 1 differ from typical learning disabled
clinic attenders: nonverbal learning disabilities revisited. Dev
Neuropsychol 2000, 17:2947.
20. Stine SB, Adams WV: Learning problems in neurofibromatosis patients. Clin Orthop 1989, 245:4348.
21. Zoller ME, Rembeck B, Backman L: Neuropsychological
deficits in adults with neurofibromatosis type 1. Acta Neurol
Scand 1997, 95:225232.
22. American Psychiatric Association: Diagnostic and statistical manual of mental disorders: DSM-IV, edn 4. Washington, DC: American Psychiatric Association; 1996:7885.
23. Kayl AE, Moore BD, Slopis JM, et al.: Quantitative morphology
of the corpus callosum in children with neurofibromatosis
and attention-deficit hyperactivity disorder. J Child Neurol
2000, 15:9096.
24. Mautner VF, Kluwe L, Thakker SD, Leark RA: Treatment of
ADHD in neurofibromatosis type 1. Dev Med Child Neurol
2002, 44:164170.
First systematic evaluation of the effectiveness of stimulant medication in children with neurofibromatosis type 1 and attention deficit
hyperactivity disorder.
25. Koth CW, Cutting LE, Denckla MB: The association of neurofibromatosis type 1 and attention deficit hyperactivity disorder. Child Neuropsychol 2000, 6:185194.
26. Ferner RE, Hughes RA, Weinman J: Intellectual impairment in
neurofibromatosis 1. J Neurol Sci 1996, 138:125133.
27. DiCapua M, Lispi ML, Giannotti A, et al.: Neurofibromatosis
type 1 presenting with hand dystonia. J Child Neurol
2001, 16:606608.
28. DiPaola DP, Zimmerman RA, Rorke LB, et al.: Neurofibromatosis type 1: pathologic substrate of high-signal intensity foci
in the brain. Radiology 1995, 195:721724.
29. Dunn DW, Roos KL: Magnetic resonance imaging evaluation
of learning disabilities and incoordination in neurofibromatosis type 1. Neurofibromatosis 1989, 2:15.
30. Duffner PK, Cohe ME, Seidel FG, Shucard DW: The significance of MRI abnormalities in children with neurofibromatosis. Neurology 1989, 39:373378.
31. Ferner RE, Chaudhuri R, Bingham J, et al.: MRI in neurofibromatosis 1. The nature and evolution of increased intensity T2
weighted lesions and their relationship to intellectual
impairment. J Neurol Neurosurg Psychiatry 1993, 56:492495.
32. Denckla MB, Hofman K, Mazzocco MM, et al.: Relationship
between T2-weighted hyperintensities (Unidentified bright
objects) and lower IQs in children with neurofibromatosis-1.
Am J Med Genet 1996, 67:98102.
33. Legius E, Descheemaeker MJ, Steyaert J, et al.: Neurofibromatosis type 1 in childhood: correlation of MRI findings with
intelligence. J Neurol Neurosurg Psychiatry 1995, 59:638640.
34. Said SM, Yeh TL, Greenwood RS, et al.: MRI morphometric
analysis and neuropsychological function in patients with
neurofibromatosis. Neuroreport 1996, 7:19411944.

35.

Cutting LE, Koth CW, Burnette CP, et al.: Relationship of

cognitive functioning, whole brain volumes, and T2weighted hyperintensities in neurofibromatosis-1. J Child
Neurol 2000, 15:157160.
36. Billingsley RL, Schrimsher GW, Jackson EF, et al.: Significance
of planum temporale and planum parietale morphologic
features in neurofibromatosis type 1. Arch Neurol
2002, 59:616622.
First article to explore relationship of sylvian fissure morphology to
cognition in neurofibromatosis type 1.
37. Cutting LE, Choe Y, Abrams MT, et al.: Gray white, and lobar
brain volumes in neurofibromatosis type 1 with and without
attention deficit hyperactivity disorder (ADHD). Neurology
2000, 54(suppl 3):A318.
38. Marchuk DA, Saulino AM, Tavakkol R, et al.: cDNA cloning of
the type 1 neurofibromatosis gene: Complete sequence of
the NF1 gene product. Genomics 1991, 11:931940.
39. Daston MM, Scrable H, Nordlund M, et al.: The protein product of the neurofibromatosis type 1 gene is expressed at
highest abundance in neurons, Schwann cells, and oligodendrocytes. Neuron 1992, 8:415428.
40. Weeber EJ, Sweatt JD: Molecular neurobiology of human
cognition. Neuron 2002, 33:845848.
Summary of current understanding of molecular neurobiology of
human cognition.
41. Martin GA, Viskochil D, Bollag G, et al.: The GAP-related
domain of the neurofibromatosis type 1 gene product interacts with ras p21. Cell 1990, 63:843849.
42. Guo HF, The I, Hannan F, et al.: Requirement of Drosphila
NF1 for activation of adenylyl cyclase by PACAP38-like neuropeptides. Science 1997, 276:795798.
43. Tong J, Hannan F, Zhu Y, et al.: Neurofibromin regulates G
protein-stimulated adenylyl cyclase activity. Nat Neurosci
2002, 5:9596.
44. Li C, Cheng Y, Gutmann DA, Mangoura D: Differential localization of the neurofibromatosis 1 (NF1) gene product, neurofibromin, with the F-actin or microtubule cytoskeleton
during differentiation of telencephalic neurons. Dev Brain Res
2001, 130:231248.
45. Bollag G, McCormick F: Regulators and effectors of ras proteins. Ann Rev Cell Biol 1991, 7:601632.
46. Reed N, Gutmann DH: Tumorigenesis in neurofibromatosis:
new insights and potential therapies. Trends Mol Med
2001, 7:157162.
47. Silva AJ, Frankland PW, Marowitz Z, et al.: A mouse model for
the learning and memory deficits associated with neurofibromatosis type 1. Nature Genet 1997, 15:281284.
48. Costa RM, Yang T, Huynh DP, et al.: Learning deficits, but normal development and tumor predisposition, in mice lacking
exon 23a of NF1. Nature Genet 2001, 27:399405.
49. Costa RM, Federov NB, Kogan JH, et al.: Mechanisms for the
learning deficits in a mouse model of neurofibromatosis
type 1. Nature 2002, 415:526530.
Novel work demonstrating ability to rescue learning deficits both
genetically and pharmacologically in neurofibromatosis type 1 mice.
50. Guo HF, Tong J, Hannan F, et al.: A neurofibromatosis-1 regulated pathway is required for learning in Drosophila. Nature
2000, 403:895898.