Pathology

The embryology relevant to neuronal proliferation and microcephaly is reviewed in the following chapter describing malformations of cortical development (Chapter 26). The pathological changes described in different types of MIC are diverse, which is not surprising, given the large number of associated conditions. Here we will confine our comments to severe congenital microcephaly. The macroscopic changes described in most pathological reports are subtle, consisting of very small cerebral volume, normal or minimally altered pattern of convolutions, and normal size of the third and lateral ventricles [Robain and Lyon, 1972]. However, our brain imaging experience shows that this is not quite true, as, in many forms, the frontal lobes are disproportionately small, and the number and complexity of the gyri and the depth of sulci are generally reduced.

The microscopic changes, especially those involving the cerebral cortex, are heterogeneous. In one group, the cortex has normal thickness and lamination, but the number of neurons in the brain is dramatically reduced. We suppose these to be the less severely affected individuals, although the available data are not clear on this point. In probably several other types of MIC, the cortex appears abnormally thin, presumably resulting from premature exhaustion of the germinal zone [Barkovich et al., 1992; Evrard et al., 1989].

In the latter, abnormalities of cellular architecture predominate in the first two layers of the cortex, referred to as “type I familial MIC” by Robain [Robain and Lyon, 1972]. Layer two is almost devoid of granule neurons, and may be fragmented into small nests (sometimes called “glomeruli”) or small columns that protrude into the molecular layer. In a few individuals, the vertical bands of neurons arising in layer two cross the molecular layer to protrude into the meninges. Neurons may be seen in the molecular layer, either as scattered large pyramidal or stellate neurons, or as persistence of a fetal monolayer of granule neurons found just beneath the pia. The lower cortical layers were less affected, but with abnormal distribution of cells in some areas. In some brains, persistence of fetal wavy or “combed” monocellular bands in the middle of the cortex has been seen [Robain and Lyon, 1972]. In these types of MIC, the cerebellum is typically small but proportionate to the reduced size of the cerebrum or relatively larger.

Severe congenital MIC has been observed in combination with several other types of brain malformations, including holoprosencephaly, disproportionate brainstem and cerebellar hypoplasia, true lissencephaly with widespread malformation of neuronal migration, diffuse periventricular nodular heterotopia, and diffuse polymicrogyria (Table 25-2).

Table 25-2 Severe Congenital Microcephaly Types by Imaging or Pathology

MIC, microcephaly; NOS, not otherwise specified.

The authors have seen agenesis of the corpus callosum in most different types of MIC, and suspect that, in most, it is a nonspecific feature of slowed brain growth. (The growing cerebral hemispheres must be closely enough apposed for the precallosal sling to cross the gap, which requires growth.) We have therefore not included MIC and agenesis as a classification in its own right at this time, although this may need to be added in the future.

Brain Imaging

In most patients with primary MIC, brain imaging reveals characteristic abnormalities that we designated “microcephaly with simplified gyral pattern” [Barkovich et al., 1998; Dobyns and Barkovich, 1999]. This pattern consists of a reduced number of gyri separated by abnormally shallow sulci. Common associated abnormalities include foreshortened frontal lobes, mildly enlarged lateral ventricles, and sometimes a thin corpus callosum or even partial agenesis of the corpus callosum.

While interpretation of brain imaging studies in MIC would seem to be straightforward, this has proved challenging in practice, primarily for severe congenital MIC. First, scans of children with MIC are often interpreted as normal, other than for the small size, if this is recognized on the imaging study, but close inspection will show the features noted above. While these changes can be subtle, they are not normal. Further, brain imaging in individuals with more severe forms of MIC may show fewer convolutions, with some broader than 2 cm, leading to interpretation as “pachygyria.” But imaging in the large majority of these patients shows a normal or an especially thin cortex, while true lissencephaly (agyria and pachygyria) is always associated with an abnormally thick cortex. Clinicians often respond to such reports by ordering tests that are appropriate for children with true lissencephaly, which are always negative. With the rare exception of mutations of TUBA1A, no genes associated with microlissencephaly (MLIS) have been reported.

While the first several genes associated with severe congenital MIC were associated with nonspecific brain imaging patterns (as described above), several recently identified MIC genes are associated with recognizable patterns of abnormalities. Focusing on children with severe congenital MIC, the authors recently reviewed brain imaging in approximately 250 children with MIC, most of whom (230 of 247) had MIC without associated somatic anomalies [Basel-Vanagaite et al., 2010]. Among this group of patients, four relatively common brain imaging patterns were found, which involved abnormalities in the gyral pattern, size of extra-axial space, and relative size of the brainstem and cerebellum in comparison to the cerebrum. The four groups were:

Examples are shown in Figure 25-1. Rare forms of severe MIC are associated with additional brain malformations, as listed in Table 25-2.

Fig. 25-1 Representative magnetic resonance imaging from each microcephaly subgroup.

A–D, Microcephaly with simplified gyral pattern only. E–F, Microcephaly with simplified gyri and pontocerebellar hypoplasia. G–H, Microcephaly with simplified gyri, enlarged extra-axial space, and proportionate cerebellum. I–J, Microcephaly with simplified gyri, enlarged extra-axial space, and marked pontocerebellar hypoplasia. K–L, Normal brain magnetic resonance images for comparison purposes. Asterisks indicate enlarged extra-axial space.

(Adapted from Basel-Vanagaite et al., Familial hydrocephalus with normal cognition and distinctive features, 2010.)

Clinical Features

The clinical manifestations associated with MIC are remarkably heterogeneous. In most individuals with severe congenital MIC, examination reveals obvious small head size, often with a low, sloping forehead and a flat occiput. The face and ears are normal, but because of the small head size, may appear disproportionately large. Cognitive impairment is moderate in some types, but severe to profound in others. In moderately affected patients, hyperactivity may dominate the patient’s behavior, while tone is typically normal. In severely affected children, spasticity and epilepsy predominate. In children with milder forms of MIC, a variety of subtle dysmorphic features may be present and may be helpful in identifying a specific syndrome causing MIC. Features of some of these syndromes are outlined in the different tables in this chapter.

For children with the most common, relatively high-functioning forms of primary MIC, survival far into adult life is typical. For more severely handicapped children unable to walk or feed by mouth, the mortality rate is higher, with survival often limited to 10–20 years, although no formal studies have been done. Children with MIC and other severe brain malformations, especially those with cortical malformations such as lissencephaly, heterotopias, and polymicrogyria, are likely to have much shorter survival.

In general, all forms of MIC are associated with below-average intelligence [Dolk, 1991; Nelson and Deutschberger, 1970]. However, mild MIC with OFC between −2 and −3 SD is not inevitably linked with mental retardation; 7.5 percent of a large group of microcephalic children had normal intelligence [Martin, 1970; Sells, 1977]. However, some patients with mild MIC have severe or profound mental retardation. Their intellectual disability may be partly explained by associated brain abnormalities, whether developmental or destructive, as brain imaging frequently reveals additional abnormalities [Sugimoto et al., 1993].

Several coexistent conditions, such as varying degrees of cognitive impairment, epilepsy, cerebral palsy, and ophthalmological and audiological disorders, occur commonly in children with microcephaly and are reviewed in the sections below.

Etiology

Mild MIC, which we define as −2 to −3 SD, has been associated with a variety of maternal and other prenatal disorders, prenatal and postnatal brain injuries, familial forms, chromosome disorders, and numerous syndromes with either prenatal- or postnatal-onset MIC. Here we can review only a small selection of the more common causes.

Genetics

At least five loci for primary microcephaly with mild or moderate phenotype have been mapped to 1q31, 8p22-pter, 9q34, 15q, and 19q13.1–q13.2, and two of these loci (MCPH1 and ASPM) have now been cloned. Two genes causing severe microcephaly with severe phenotype have also been identified, those being the causal genes for Amish lethal microcephaly and microcephaly with periventricular nodular heterotopia. In Seckel’s syndrome, defects in DNA repair were suggested by chromosome instability with exposure to mitomycin C [Abou-Zahr et al., 1999; Bobabilla-Morales et al., 2003]. Subsequently, at least three loci for Seckel’s syndrome have been confirmed, mapped to chromosomes 3q22.1–q24 and 18p11.31–q11.2, and at least one additional unknown locus [Faivre et al., 2002]. Mutations of the ATR gene in 3q2 were recently identified [Alderton et al., 2004; O’Driscoll and Jeggo, 2003]. All of these genes are listed in Table 25-6, with references. Of interest, the two genes associated with microcephaly and mild–moderate phenotype both reveal an evolutionary signature of rapid evolution [Evans et al., 2004a, b].

Antenatal Diagnosis

Microcephaly can often, but not always, be diagnosed by second-trimester fetal ultrasonography [Bromley and Benacerraf, 1995]. This likely is due to variable onset of the deceleration in head growth. When this occurs early, as it often does for severe microcephaly, ultrasound examination should be able to detect the abnormality, but not when it begins in the late second or third trimester.

Genetic Counseling

Some older references cite a 6 percent risk of a family’s having a second microcephalic child, but these sources do not consistently address severity of the microcephaly. This percentage may be useful for mild and borderline microcephaly with birth occipitofrontal circumference between −2 and −3 SD below the mean. On the basis of findings in many families with two or more affected siblings with primary microcephaly and other forms of severe microcephaly with birth occipitofrontal circumference below −3 SD, counseling for autosomal-recessive inheritance is appropriate in this group. Thus, most forms of severe congenital MIC (with or without intrauterine growth retardation) are genetic, most if not all having autosomal-recessive inheritance. Disorders associated with postnatal MIC are much more heterogeneous, with examples of autosomal-dominant (familial or sporadic), autosomal-recessive, and X-linked inheritance (see Table 25-6).

Chromosome Disorders

Many chromosome disorders, including the common trisomies (trisomies 13, 18, and 21), and many structural rearrangements such as cri du chat syndrome (deletion 5p15), are associated with MIC. For most individuals, this presents as a mild congenital form that often is followed by more severe postnatal MIC.

An OMIM search lists more than 400 syndromes with microcephaly, making this an unhelpful search term. Some of the better-known disorders include Angelman’s, Cornelia de Lange (Brachmann–de Lange), and Dubowitz’s syndromes [Opitz and Holt, 1990].

With potentially hundreds of causes of microcephaly, including prenatal and postnatal onset, as well as genetic and acquired etiologies, diagnostic evaluations may be complex. Investigation of patients with microcephaly includes evaluation for prenatal exposure to teratogens, especially alcohol, drugs, and isotretinoin (a vitamin A analog), and assessment of the family history, birth history, and associated malformations. Laboratory studies should include titers for toxoplasmosis, syphilis, rubella virus, cytomegalovirus, and herpes simplex viruses; neuroimaging [Sugimoto et al., 1993]; evaluation for maternal and childhood metabolic disorders; and genetic testing, including chromosome analysis and testing for small deletions or duplications, which currently is performed by fluorescence in situ hybridization with subtelomeric probes [Knight et al., 2000]. Algorithms for the evaluation of the infant and child with congenital (Figure 25-2) and postnatal (Figure 25-3) microcephaly have recently been published and serve as a generalized approach to the diagnostic evaluation [Ashwal et al., 2009].

Fig. 25-2 Algorithm for the diagnostic evaluation of the infant or child with congenital microcephaly.

(Adapted from Ashwal S, et al. Practice parameter: Evaluation of the child with microcephaly [an evidence-based review]: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society, Neurology 73:887–897, 2009.)

Fig. 25-3 Algorithm for the diagnostic evaluation of the infant or child with postnatal-onset microcephaly.

(Adapted from Ashwal S, et al. Practice parameter: Evaluation of the child with microcephaly [an evidence-based review]: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society. Neurology 73:887–897, 2009.)

Pathology and Pathogenesis

Numerous animal models of syndromic and nonsyndromic megalencephaly display neuronal and glial hypertrophy. Pten (phosphatase and tensin homolog on chromosome ten) mutant mice were found to develop macrocephaly and behavioral abnormalities reminiscent of human autistic spectrum disorder, such as reduced social activity, increased anxiety, and sporadic seizures [Kwon et al., 2001; Kwon et al., 2006; Ogawa et al., 2007], closely resembling the human phenotype of PTEN-related disorders that are described later in this chapter. At the cellular level, in vivo effects of loss of Pten include loss of neuronal polarity, neuronal hypertrophy, and, in one study, increased astrocyte proliferation and hypertrophy [Kwon et al., 2001; Fraser et al., 2004]. Increasing attention has been paid to the role of the mammalian target of rapamycin (mTOR), a serine/threonine kinase that has well-known functions in regulation of cellular proliferation and growth, a crucial role in neuronal development and synaptic plasticity [Jaworski and Sheng, 2006], and a contribution to Pten-mediated growth regulation in the mammalian nervous system. mTOR inhibition reversed neuronal hypertrophy in Pten-deficient mice and also resulted in amelioration of a subset of Pten-associated abnormal behaviors, thereby substantiating evidence that the mTOR pathway downstream of PTEN is critical for its complex phenotype [Kwon et al., 2003; Zhou et al., 2009].

Loss of Tsc1 and Tsc2, two downstream negative regulators of the mTOR pathway [Inoki et al., 2002; Manning et al., 2002; Potter et al., 2002], has been shown to cause neuronal hypertrophy in vitro and in vivo [Jaworski et al., 2005; Tavazoie et al., 2005; Meikle et al., 2007], supporting a role of TSC1 and TSC2 in neuronal growth regulation and synaptic function. Interruptions of TSC1 and TSC2 cause tuberous sclerosis complex, known to be associated with megalencephaly, hemimegalencephaly, and focal megalencephaly [Choi et al., 2008]. Zhou et al. suggested that there is a common signal transduction pathway potentially responsible for the autism-like symptoms in individuals bearing TSC1/2 and/or PTEN mutations, and proposed that mTOR inhibitors are potential therapeutic agents for this subset of patients [Meikle et al., 2008; Zhou et al., 2009]. The Nf1 knockout mouse was found to have increased neuroglial progenitor/stem cell (NSC) proliferation and gliogenesis in the brainstem, also driven by mTOR-mediated activation [Lee et al., 2010].

Other animal models of megalencephaly include mouse mutants with loss-of-function mutations in genes regulating programmed cell death, or apoptosis, such as Caspase-3, Caspase-9, and Apaf-1, which were found to have gross brain malformations and neuronal hyperplasia. However, these mutations, when germline, are embryonically lethal [Kuida et al., 1996, 1998; Cecconi et al., 1998; Yoshida et al., 1998; Hakem et al., 1998; Marks and Berg, 1999]. Transgenic mice overexpressing insulin-like growth factor (IGF)-I exhibit brain overgrowth characterized by increased numbers of neurons and oligodendrocytes, as well as excessive myelin formation [Carson et al., 1993; Donahue et al., 1996; Petersson et al., 1999; D’Ercole et al., 2002]. IGF-1 stimulates:

As a result of these events, brain growth is increased with IGF-I overexpression and reduced with decreased IGF-I signaling. Although much less information is available in humans, individuals with IGF-I gene deletions or mutations that result in severe deficits in IGF-1 expression are microcephalic and mentally retarded [Walenkamp and Wit, 2007]. Little evidence supporting comparable actions for IGF-II is available.

The CD81 null mouse has a markedly increased brain size (up to 30 percent larger) due to an increased number of astrocytes and microglia throughout the brain, possibly through regulation of cell proliferation by a contact inhibition-dependent mechanism. CD81 is a member of the tetraspanin family of small membrane proteins associated with the regulation of cell migration and mitotic activity [Geisert et al., 2002]. In yet another animal model, transgenic mice expressing a stabilized β-catenin in neural precursors develop enlarged brains with increased cerebral cortical surface area and folds resembling sulci and gyri of higher mammals [Chenn and Walsh, 2002, 2003]. Brains from these animals have enlarged lateral ventricles lined with neuroepithelial precursor cells that are derived from an expanded precursor population. Compared with the wild type of precursors, a greater proportion of transgenic precursors re-enter the cell cycle after mitosis, which suggests that β-catenin regulates cerebral cortical size by controlling the generation of neural precursor cells.

Among the few models with postnatal progressive megalencephaly are the epileptic megalencephaly BALB/cByJ-Kv1.1^mceph/mceph (called mceph/mceph) mice [Donahue et al., 1996] and the epileptic (epi/epi) chicken [George et al., 1990a]. The mceph/mceph mice carry a spontaneous germline mutation in a gene encoding a potassium ion channel subunit. This mutation makes the channel protein, Kv1.1, nonfunctional and causes complex partial epilepsy with the limbic system as the major focus (temporal lobe epilepsy [TLE]); interestingly, in parallel to progressive epileptic behavior, the mceph/mceph brains show progressive overgrowth, in the absence of other structural brain abnormalities. This excessive brain enlargement is restricted to the hippocampus and ventral cortical structures, including the piriform/entorhinal cortex and amygdala, whereas the thalamus, olfactory bulb, and cerebellum have wild-type sizes. The volume increase in the mceph/mceph hippocampus is due to a doubling of the number of neurons and astrocytes. In humans, Kv1.1 mutations, where only one amino acid is changed, have been found in patients with epilepsy or episodic ataxia type 1 (EA1). From extensive studies of the mceph/mceph mouse, it has been hypothesized that some human idiopathic megalencephalies with severe early-onset seizures are caused by such severe ion channelopathies [Almgren et al., 2008].

Clinical Features

Etiology

The most significant macrocephaly (and/or megalencephaly) syndromes are listed in Tables 25-9 and 25-10, with a brief overview of their clinical features, MRI findings, and genetic bases. These include classic overgrowth syndromes, such as Sotos’, Weaver’s, and Simpson–Golabi–Behmel syndromes; PTEN-related disorders, such as Cowden’s and Bannayan–Riley–Ruvalcaba syndromes; the macrocephaly-capillary malformation (M-CM) syndrome (previously termed macrocephaly cutis marmorata telangiectatica congenita, or CMTC); and skeletal dysplasias, such as achondroplasia and thanatophoric dysplasia, as well as a number of chromosomal disorders. By far, the majority of these disorders are inherited as an autosomal-dominant trait. Their clinical features and neurodevelopmental outcome are quite variable and dependent on the ensuing neuronal dysfunction caused by the specific underlying disorder. The most notable megalencephaly/macrocephaly disorders are discussed briefly below.

Overgrowth Syndromes

Macrocephaly frequently occurs in conjunction with body overgrowth (height and weight >2 SD above the mean for age), as in Sotos’, Weaver’s, and Simpson–Golabi–Behmel syndromes. Many of these overgrowth disorders are characterized by excessive growth in fetal life and infancy, with subsequent decline in growth rate and normalization of growth in adulthood. Partial (or focal) and unilateral overgrowth (or hemihypertrophy) is occasionally seen in other MEG or MAC syndromes, such as M-CM syndrome, PTEN-related disorders, and some chromosomal disorders such as Pallister–Killian syndrome. Children who are macrocephalic at birth may become normocephalic or relatively microcephalic when older, if body overgrowth supersedes brain growth, as typically occurs in Beckwith–Wiedemann syndrome.

Sotos’ syndrome is an autosomal-dominant disorder due to mutations or deletions of NSD1 (nuclear receptor-binding SET domain protein-1). Macrocephaly is usually present at all ages in more than 90 percent of children and is considered to be a cardinal feature [Agwu et al., 1999; Rio et al., 2003; Tatton-Brown et al., 2005]. In some series, macrocephaly was present at birth in 50 percent of children, with birth OFCs as high as +4 above the mean, and later OFCs ranging between +2 and +7 SD. Most patients have a nonprogressive neurologic dysfunction characterized by clumsiness and poor coordination [Cole and Hughes, 1994]. Delays in expressive language and motor development during infancy are particularly common and, in some instances, may be followed by attainment of normal or near-normal intelligence. Several patients with Sotos’ syndrome and autistic features have been reported [Morrow et al., 1990; Battaglia and Carey, 2006]. Seizures and tone abnormalities are occasionally present [Cohen, 1989, 1999; Cole and Hughes, 1990, 1994]. Brain MRI abnormalities present in patients with Sotos’ syndrome and an NSD1 mutation include enlarged extra-axial fluid and lateral ventricles in 70 percent and 60 percent of patients, respectively, and it has been suggested that these increased CSF spaces are primarily responsible for macrocephaly in Sotos’ syndrome, rather than true megalencephaly [Schaefer et al., 1997]. Between 80 and 90 percent of patients have a demonstrable NSD1 abnormality. NSD1 is involved in an intricate regulatory network of genes that appear to have a concerted role in various processes, including cell growth and tumorigenesis [Lucio-Eterovic et al., 2010].

NSD1 mutations have also been found in a significant proportion of patients with Weaver’s syndrome, a rarer overgrowth disorder characterized by macrocephaly, dysmorphic facial features (especially prominent hypertelorism), metaphyseal flaring of the femurs, camptodactyly, deep-set nails, and hoarse, low-pitched cry. Therefore, significant clinical and genetic overlap exists between these two disorders of macrocephaly and overgrowth [Proud et al., 1998; Rio et al., 2003; Cecconi et al., 2005].

Simpson–Golabi–Behmel syndrome (SGBS) is an X-linked complex congenital overgrowth syndrome characterized by macroglossia, macrosomia, renal and skeletal abnormalities, and an increased risk of embryonal tumors. Macrocephaly is often congenital. Patients may have hypotonia and mild developmental delay, although most have normal intelligence [Neri et al., 1998]. Most cases of SGBS are due to mutations or deletions of the glypican-3 (GPC3) gene at Xq26, a member of a multigene family encoding at least six distinct glycosylphosphatidylinositol-linked cell-surface heparan sulfate proteoglycans (HSPGs); these act as co-receptors for multiple families of growth factors that have been shown to regulate cell proliferation, differentiation, and patterning, including that of the brain. In support of the glypicans’ role in development, mice with null mutations in glypican-1 (Gpc1) have a severely reduced brain size and an abnormally small-sized cerebellum. Therefore, Gpc1 may have a role in early neurogenesis, possibly through regulation of fibroblast growth factor (fgf) signaling [Jen et al., 2009].

SGBS type 2 is an X-linked mental retardation syndrome with macrocephaly (OFCs +2 to +6 SD above the mean) and ciliary dysfunction, manifesting as recurrent respiratory tract infections, with abnormal functional studies of the respiratory cilia. Recently, a family with this syndrome co-segregating with a frameshift mutation in the oral-facial-digital type 1 (OFD1) gene was reported [Budny et al., 2006].

RASopathies

The Ras/mitogen-activated protein kinase (MAPK) pathway is essential in the regulation of the cell cycle, cell differentiation, growth, and cell senescence, each of which is critical to normal development. The “RASopathies” are a class of developmental disorders caused by germline mutations in genes that encode protein components of the Ras/MAPK pathway, which result in dysregulation of the pathway and profoundly deleterious effects on development. These disorders include neurofibromatosis type 1 (NF1), Costello’s syndrome (HRAS), cardiofaciocutaneous (CFC) syndrome (KRAS, BRAF, and MEK1), and Noonan’s syndrome (PTPN11, KRAS, and SOS1), among others. Neurofibromatosis type 1 (NF1) is a disorder of true megalencephaly, whereas Noonan’s, Costello’s, and CFC syndromes have a high incidence of relative macrocephaly and ventriculomegaly.

NF1 shares features of other overgrowth syndromes, such as the presence of macrocephaly, various types of tumors, and, occasionally, hemihyperplasia of a limb or digit, despite an increased incidence of short stature. Macrocephaly in the absence of hydrocephalus occurs in 50 percent of individuals with NF1 [Tonsgard, 2006]. Quantitative MRI studies have demonstrated the presence of true megalencephaly, largely secondary to increased white-matter volume [Bale et al., 1991; Said et al., 1996; Steen et al., 2001; Cutting et al., 2002]. Learning disabilities have been reported in up to 70 percent of individuals, and 3 percent have severe developmental delay. Their neurocognitive profile may also include easy distractibility, impulsiveness, and deficient visual-motor coordination. Seizures occur in approximately 6–7 percent of patients. Frank hydrocephalus with aqueductal stenosis, as well as asymptomatic ventricular dilatation, has been observed in approximately 4 percent of patients. NF1 is a tumor suppressor gene, expressed in neurons and glial cells, which encodes neurofibromin, one of the earliest identified regulators of the RAS-MAPK pathway; it thus has important roles in cellular proliferation and differentiation [Daston et al., 1992; Nordlund et al., 1993; Cichowski and Jacks, 2001].

Recently, a dominant condition that overlaps with NF1 clinically (with macrocephaly, café au lait lesions, and axillary freckling) has been described in association with heterozygous mutations in SPRED1, a member of the SPROUTY/SPREAD family of proteins that are also regulators of RAS–RAF interaction and MAPK signaling [Brems et al., 2007] (see Legius’ syndrome in Tables 25-9 and 25-10).

Costello’s syndrome is a unique combination of failure to thrive, cardiac abnormalities, and a predisposition to papillomata and malignant tumors. In a systematic review of 28 patients, absolute or relative macrocephaly was found in 100 percent of patients, and, more specifically an evolving megalencephaly and cerebellar enlargement, overlapping with M-CM syndrome [Gripp et al., 2010]. Neurologic abnormalities include developmental delay/mental retardation, nystagmus, and hypotonia [Quezada and Gripp, 2007; Gripp and Lin, 2009].

Macrocephaly-Capillary Malformation Syndrome

The macrocephaly-capillary malformation (MCAP) syndrome is a distinct syndrome characterized by megalencephaly, vascular malformations (most often cutis marmorata), hemihypertrophy, digit anomalies, and skin and connective tissue laxity. More than 100 patients with M-CM syndrome have been reported [Clayton-Smith et al., 1997; Moore et al., 1997; Vogels et al., 1998; Thong et al., 1999; Franceschini et al., 2000; Robertson et al., 2000; Giuliano et al., 2004; Lapunzina et al., 2004; Canham and Holder, 2008], and its neuroimaging findings were reviewed by Garavelli et al. [2005] and Conway et al. [2007]. The megalencephaly and perisylvian polymicrogyria with postaxial polydactyly and hydrocephalus (MPPH) syndrome is a more recently described syndrome in an initial cohort of five patients [Mirzaa et al., 2004], and four subsequent single cases [Colombani et al., 2006; Garavelli et al., 2007; Tohyama et al., 2007; Pisano et al., 2008]. Since then, a marked increase in ascertainment of patients with overlapping features of both syndromes has been witnessed, and it is proposed that they represent a single megalencephaly syndrome [Gripp et al., 2009; unpublished data]. MEG is most often congenital, with OFCs ranging from +2 to +4 SD above the mean at birth, and reaching up to +8 SD later in life. Variable degrees of developmental delay, hypotonia, and seizures occur. Vascular anomalies are a characteristic and defining feature and most commonly consist of cutis marmorata, the cutaneous marbled appearance frequently seen in Caucasian newborns that tends to fade with time but often persists. Other vascular anomalies include a midline nevus flammeus, various types of hemangiomas in any location, vascular rings, and telangiectasias. Digit anomalies include the common 2–3 toe syndactyly (>25 percent syndactyly), 2–3–4 finger syndactyly, and postaxial polydactyly. Common MRI abnormalities (Figure 25-4A) include diffuse megalencephaly that is symmetric or mildly asymmetric, a very high rate of hydrocephalus that is often shunted, or ventriculomegaly, progressive posterior fossa crowding with cerebellar tonsillar herniation that may require decompression, polymicrogyria that is by far bilateral perisylvian in distribution, and white-matter abnormalities. A distinct subset of patients has a very thick (or mega-) corpus callosum [Conway et al., 2007; unpublished data]. Serial neuroimaging has demonstrated that, despite shunting procedures, OFCs continue to follow an accelerated growth rate, thereby demonstrating the presence of true megalencephaly. All reported cases to date appear sporadic.

Fig. 25-4 Subtypes of megalencephaly and hemimegalencephaly.

Right parasagittal (left column, except midsagittal in D), left parasagittal (middle column), and axial (right column) magnetic resonance images from four patients with megalencephaly (MEG) or hemimegalencephaly (HMEG) variants. A, The top row images depict symmetric MEG and perisylvian polymicrogyria with normal white matter. The patient was a female with the originally described “megalencephaly polymicrogyria polydactyly hydrocephalus” (MPPH) syndrome [Mirzaa et al., 2004]. The symmetry and normal white matter distinguish this malformation from HMEG. B, The second row images show partial HMEG, with enlargement of the posterior frontal, temporal, and parietal lobes on the right. The abnormal white matter typical of HMEG is seen circling the back of the right lateral ventricle. C, The third row images demonstrate severe HMEG involving the entire right hemisphere, but sparing the left. The central and deep white matter has diffusely bright signal, sparing only the superficial U fibers. D, The bottom row images show a very rare malformation consisting of bilateral HMEG that is more severe on the left side. The patient survived only a few months.

(Courtesy of Dr. William B Dobyns, University of Washington and Principal Investigator, Center for Integrative Brain Research, Seattle Children’s Research Institute, Seattle, WA.)

Perhaps as a severe variant of this syndrome, Gohlich-Ratmann et al. reported three sporadic cases with congenital megalencephaly, a greatly hypertrophied corpus callosum, and complete lack of motor development [Gohlich-Ratmann et al., 1998]. Cranial MRI demonstrated bilateral and symmetric megalencephaly and polymicrogyria. Two cases were subsequently reported with similar features, one with minimal motor development (with the ability roll sideways only) [Dagli et al., 2008; Hengst et al., 2010]. Two children from a consanguineous family were similarly reported with mega-corpus callosum, polymicrogyria, and moderate psychomotor retardation, suggestive of autosomal-recessive inheritance. These patients additionally exhibited pontine and cerebellar vermis hypoplasia [Bindu et al., 2010].

PTEN-Related Disorders

PTEN is a tumor suppressor gene, somatic mutations of which have been reported to varying degrees in multiple sporadic malignancies (such as glioblastoma multiforme, among others) [Eng, 2000, 2003]. Germline mutations of PTEN have been found in a set of disorders of macrocephaly and hamartomatous overgrowth, namely Cowden’s (CS), Bannayan–Riley–Ruvalcaba (BRRS), and Proteus’ syndromes, and in a subset of patients with a “Proteus-like” phenotype. CS and BRSS have a high degree of clinical overlap and are believed to constitute a single clinical spectrum (CS-BRRS). Macrocephaly is a prominent and progressive feature, with OFCs typically +4.5 SD or more above the mean, and reaching up to +8 SD. Hypotonia and delayed gross motor skills are common findings. Around 60 percent of patients have a mild proximal myopathy, and 25 percent have seizures. Additional features include hamartomas, lipomas, intestinal polyps, and various types of cutaneous vascular malformations. PTEN mutation carriers are at increased risk for various tumors (most notably of the breast, thyroid, and endometrium).

Proteus’ syndrome (PS) is a rare and highly variable disorder with relentless asymmetric and disproportionate overgrowth of body parts, vascular malformations, cerebriform connective tissue nevi, epidermal nevi, and dysregulated adipose tissue [Cohen and Hayden, 1979; Wiedemann et al., 1983; Cohen et al., 2002], which has been reported in association with HMEG or unilateral MEG. Given the genetic overlap between these disorders of dysregulated cellular proliferation, the term “PTEN hamartoma tumor syndrome” (PHTS) has been coined for this group of distinct conditions [Marsh et al., 1999; Eng, 2000]. PTEN mediates cell cycle arrest and/or apoptosis by negatively regulating the phosphinositide-3-kinase-Akt serine/threonine protein kinase (PI3K/Akt) pathway [Furnari et al., 1998; Li et al., 1998; Weng et al., 1999]. Accumulating evidence suggests that PTEN also regulates cell survival pathways, such as the MAPK pathway [Gu et al., 1998; Simpson and Parsons, 2001; Weng et al., 2001, 2002]. PTEN mutations have recently been identified in patients with isolated macrocephaly and autistic spectrum disorders (ASDs), and/or developmental delay, as discussed below.

Macrocephaly-Autism Syndrome

An increased rate of macrocephaly is a consistent and replicated biological finding in ASD, and appears to be the single most consistent physical characteristic of children with autism. Multiple studies of OFC in persons with ASD have shown that macrocephaly occurs more frequently than expected [Bailey et al., 1993; Bolton et al., 2001; Davidovitch et al., 1996; Woodhouse et al., 1996; Lainhart et al., 1997; Stevenson et al., 1997; Fombonne et al., 1999; Fidler et al., 2000; Miles et al., 2000; Aylward et al., 2002; Gillberg and De Souza, 2002; Deutsch and Joseph, 2003; Dementieva et al., 2005; Lainhart et al., 2006; Courchesne et al., 2010]. These studies show, on average, a rate of macrocephaly of 20 percent in patients with ASD [Fombonne et al., 1999]. Neuroimaging studies of autism have found increased mean total brain volume in children by 2–4 years of age [Courchesne et al., 2001; Sparks et al., 2002; Hazlett et al., 2005]. Furthermore, postmortem studies show increased brain weight in children with autism, with frank megalencephaly in some [Bailey et al., 1998; Kemper and Bauman, 1998]. Causes of macrocephaly other than increased brain volume are rarely found in patients with idiopathic autism [Lainhart et al., 1997; Stevenson et al., 1997; Bailey et al., 1998; Bigler et al., 2003]. Given these data, this common association has been termed the “macrocephaly-autism syndrome.”

Butler and colleagues [2005] published the first report directly linking PTEN and isolated ASDs, identifying mutations in 3 of 18 individuals. These 3 subjects were boys, 2 of whom had the largest OFCs in the cohort, of +7 SD and +8 SD above the mean [Butler et al., 2005]. A similar study of 71 patients with isolated ASD (57 of whom met the Diagnostic and Statistical Manual of Mental Disorders [DSM]-IV criteria for autism) identified PTEN mutations in 2 out of 16 tested individuals who also had macrocephaly [Herman et al., 2007]. Using individuals with ASD and macrocephaly drawn from the Paris Autism Research International Sibpair (PARIS) study, the Autism Genetic Research Exchange [AGRE], and separately recruited patients, Buxbaum et al. found a PTEN mutation in 1 of 88 subjects tested [Buxbaum et al., 2007].

Subsequently, PTEN mutations were found in 5 of 60 (8.3 percent) individuals with macrocephaly and ASD, and 6 of 49 (12.2 percent) individuals with macrocephaly and developmental delay (DD)/mental retardation (MR) [Varga et al., 2009]; these results were extended by a cohort study whereby PTEN mutations were found in 7 of 99 (7.1 percent) individuals with MAC/ASD and 8 of 100 (8 percent) of individuals with MAC/DD [McBride et al., 2010]. Therefore, the estimated PTEN mutation frequency in macrocephaly-autism syndrome is approximately 20 percent overall. It is interesting to note that, from a molecular standpoint, mutations in TSC1/TSC2, NF1, or PTEN activate the mTOR/PI3K pathway and lead to syndromic ASD with tuberous sclerosis, neurofibromatosis, or macrocephaly, respectively, as mentioned above.

Disorders of Skeletal Involvement

Achondroplasia and thanatophoric dysplasia are well known to be associated with true megalencephaly from multiple reports in the early literature [Dennis et al., 1961; Cohen et al., 1967; Priestley and Lorber, 1981; Knisely, 1989]. MEG frequently coexists with mild ventriculomegaly, with or without hydrocephalus related to stenosis of the sigmoid sinus at the level of narrowed jugular foramina that rarely requires surgical intervention [Pierre-Kahn et al., 1980]. Robinow’s syndrome is a genetically heterogeneous condition characterized by mesomelic limb shortening and facial and genital anomalies. The estimated frequency of MEG (which may be congenital) is up to 64 percent in the dominant form and 25 percent in the recessive form [Mazzeu et al., 2007]. Greig’s cephalopolysyndactyly syndrome (GCPS) is a rare pleiotropic, multiple congenital anomaly syndrome characterized by the triad of polysyndactyly, hypertelorism, and macrocephaly, which is not typically associated with other central nervous system anomalies [Biesecker, 2008]. The acrocallosal syndrome resembles GCPS, with the presence of preaxial polysyndactyly and macrocephaly, but is distinguished by agenesis of the corpus callosum, and a much higher incidence of seizures and mental retardation [Johnston et al., 2003].

Chromosomal Abnormalities

Macrocephaly has been reported in several chromosomal disorders, such as trisomy 5p [Leschot and Lim, 1979; Reichenbach et al., 1999], partial duplications of 12p [Rauch et al., 1996], proximal and distal 15q duplications [Hood et al., 1986; Roggenbuck et al., 2004], deletions of 22q13 [Phelan et al., 1992; Tabolacci et al., 2005], and Pallister–Killian syndrome due to mosaic isochromosome 12p [Smigiel et al., 2008]. A number of reciprocal microdeletion and microduplication syndromes are associated with MIC/MAC, such as those involving 1q21.1 [Brunetti-Pierri et al., 2008], 2p24.3 (the locus for the MYCN gene) [Malan et al., 2010], and 5q35.3 (the NSD1 locus) [Lucio-Eterovic et al., 2010]. MEG has also been reported in two patients with Klinefelter’s syndrome. One had frank, bilateral, and symmetric MEG, in association with polymicrogyria and neuronal heterotopias [Budka, 1978]; the other had mild, focal, and unilateral MEG, in association with macrogyria [Choi et al., 1980].

Other Megalencephaly/Macrocephaly Syndromes

Relative or absolute macrocephaly occurs in some patients with Opitz–Kaveggia (or FG) syndrome, an X-linked disorder characterized by imperforate anus, broad and flat thumbs, hypotonia, moderate to severe DD, and abnormalities of the corpus callosum, and is due to mutations in MED12. Recently, an allelic disorder consisting of macrocephaly, tall, thin (or Marfanoid) habitus, and distinct facial features, called Lujan (or Lujan–Fryns) syndrome, was linked to MED12 as well.

Other rarer macrocephaly syndromes with few reports in the literature include the macrocephaly, megalocornea, motor and mental retardation (MMMM) syndrome, macrosomia, obesity, macrocephaly, and ocular abnormalities (MOMO) syndrome, and Clark–Baraitser syndrome. A single old report described a patient with ataxia telangiectasia, megalencephaly, and intestinal lymphangiectasia [Scott, 1969].