Clinical manifestations

Presentation and course

A patient with megalencephaly may present to the clinician simply with the chief complaint of asymptomatic macrocephaly, often detected on a routine well-child visit. The routine measurement of the head circumference in the primary care setting, with plotting on a standardized occipital-frontal circumference graph, is 1 of the most common means of discovering a child whose head size lies outside the expected range for the patient's age and gender (94; 38). Associated clinical findings, if present, vary widely depending on the underlying etiology and pathogenesis.

Patients with normal or large heads at birth, who experience rapid acceleration in head size, with bulging fontanelles, splaying of sutures, or other signs of increased intracranial pressure, are usually evaluated via emergent neuroimaging and diagnosed with hydrocephalus. Alternatively, individuals with anatomic megalencephaly often have a large head at birth but have normal intracranial pressure; although head size increases to above the 98th percentile within the first few months of life, the rate of growth then begins to parallel the upper curve on occipital-frontal circumference (OFC) growth curves. Some of these children have physical manifestations of neurocutaneous disorders or overgrowth syndromes. In contrast, children with metabolic megalencephaly typically have a normal head circumference at birth, with subsequent relatively rapid head growth associated with developmental delay or regression (97; 138; 38).

One of the more common forms of anatomic megalencephaly is primary or benign familial megalencephaly (which overlaps with benign familial macrocephaly). Children usually have a large head at birth and demonstrate rapid enlargement in head circumference during the first few months after birth. Thereafter, head growth usually parallels the upper percentiles of normal curves. Patients may have accelerated growth, exceeding the 98th percentile for many months, only to decelerate in growth rate, finally achieving a statistically large, but clinically insignificant, head circumference. These individuals are correctly diagnosed with “macrocephaly,” but by definition some also have “megalencephaly,” as both occipital-frontal circumference and intracranial contents are statistically larger than average.

There is overlap between benign familial megalencephaly and familial macrocephaly, such that children with large heads due to large brains or enlarged subarachnoid spaces (benign enlargement of subarachnoid spaces of infancy) both often have family members with macrocephaly and megalencephaly. Part of the confusion is that much of the data published on these conditions predates modern imaging. Furthermore the terms megalencephaly and macrocephaly have been used interchangeably in the literature to describe both the same and different conditions. Both are presumably inherited as autosomal dominant traits and, thus, are likely related or overlapping clinical conditions; children with large heads due to either large brains or large subarachnoid spaces that are otherwise normal often grow into normal adults with large heads and brains (29; 30; 44; 43; 61; 111; 41; 38).

Developmental regression or significant delays are not typical in such individuals with benign familial megalencephaly, but mild gross motor or psychomotor delays or learning disabilities may be seen (52). Static or progressive coarsening of facial features or other facial dysmorphisms, skeletal deformities (frontal bossing, short digits, widening of ribs), hepatosplenomegaly, dwarfism, gigantism, seizures, hypotonia, hyperreflexia, incoordination, diplegia, extrapyramidal signs, visual impairment, and neurocutaneous lesions are associated features that provide clues to some forms of anatomic or metabolic megalencephaly (29; 30; 45; 43; 111; 97; 138; 38).

Persons with clinically significant megalencephaly may be characterized as having 1 or more of the following:

(1) Head circumference measurements performed either at birth or later in infancy that reveal a head size more than 2 standard deviations above the mean in an otherwise normal individual. Successive measurements may demonstrate an accelerated growth rate. However, this finding alone may also be seen in clinically-insignificant megalencephaly and, thus, merely indicates that the child should be followed for abnormalities in development, rate of head growth, or neurologic condition, and symptoms of increased intracranial pressure should be evaluated.

(2) Mental or psychomotor retardation, learning difficulties, or seizures.

(3) Stigmata associated with neurocutaneous syndromes.

(4) Somatic anomalies, including excessive stature, dwarfism, abnormal facies, or other developmental malformations.

In addition, clinically-significant megalencephaly may affect 1 or both hemispheres. In unilateral forms (hemimegalencephaly), associated clinical features again vary depending on the underlying etiology. Patients may or may not have somatic hemihypertrophy, extraneural hamartomatous lesions, intractable seizures, mental retardation, hyperreflexia, hypotonia, unilateral infantile spasms, Chiari I malformation with holocord syrinx, or even unilateral alopecia (127; 30; 98; 39; 40; 43; 111; 13; 116).

Megalencephaly may be present without any pathognomonic findings of the various associated syndromes. Although some authors have emphasized the association with mental retardation and seizures (36; 74), the most common type of megalencephaly, benign familial megalencephaly, may have no associated clinical or other pathologic features (43; 111; 38).

Lorber and Priestley reviewed the records of children referred over a 20-year period for evaluation of large head without associated features. Of 510 children with occipital-frontal circumference greater than the 98th percentile, 384 had increased intracranial pressure. Of the remaining 126 with megalencephaly and normal intracranial pressure, 17 had specific syndromes associated with clinically-significant megalencephaly. The remaining 109 children were believed to have clinically-insignificant megalencephaly or “primary megalencephaly.” Ninety-six of these children had normal or above-average intelligence, and an additional 6 were found to be below average but within the normal range. Of the remaining 7 children whose testing placed them in the mentally retarded range, 2 had a strong family history of retardation, and 6 had normal CT scans. Sixty-two of these children had a parent or close relative with a large head (most commonly the child’s father), and the male to female ratio was 4:1 (75).

Other studies of large pedigrees suggest that intellect is not always affected and that many individuals with large heads often have benign familial megalencephaly or benign familial macrocephaly (101; 114; 08; 26; 100; 43; 111; 38).

Clinically-insignificant or benign/primary megalencephaly is characterized by the following criteria (30):

(1) Head circumference greater than 2 standard deviations above the mean.

(2) No clinical evidence of increased intracranial pressure.

(3) Normal developmental and neurologic examinations.

(4) Absence of any constellation of neurocutaneous stigmata and craniofacial or somatic anomalies suggesting a specific syndrome.

(5) One or both parents or siblings has a large occipital-frontal circumference (OFC) but is neurologically normal, or the enlarged OFC can be traced through several generations.

(6) Clinical monitoring at follow-up visits establishes normal development and slowing of the rate of head growth so that the growth acceleration begins to parallel the normal curve.

Many authors support DeMyer's criteria that patients with primary/benign megalencephaly have a normal neurologic exam, normal development, and at least 1 affected family member who also has normal intelligence (101; 114; 08; 26; 100). Patients may occasionally have neuroimaging findings of mildly enlarged ventricles or subarachnoid spaces. Rare individuals have demonstrated mild suture diastasis on skull films, but both neuroimaging and radiographic findings may be transient (08; 26; 100; 75; 05; 72). A study of 732 boys with primary megalencephaly showed an increased association of mental retardation with large brain size, with an odds ratio of 1.32 with 95% confidence intervals of 0.80 to 2.02 (99). Two other series showed increased rates of mental retardation, seizures, and brain malformations on autopsy, demonstrating at least that children with megalencephaly and mental retardation have higher rates of additional brain malformations (28; 72). However, in general these patients have good outcome and minor delays or neurologic abnormalities (43; 111; 38). DeMyer referred to patients with megalencephaly who had positive family histories of megalencephaly but no clear-cut syndrome and only mild motor or mental disability as “symptomatic familial anatomic megalencephaly” (30).

Children with subarachnomegaly (benign enlargement of subarachnoid spaces of infancy) likely represent a variant of benign megalencephaly and mostly have been reported to have relatively normal development. Ment and colleagues reported 17 of 18 infants had no evidence of neurologic abnormality, and the 1 child that showed abnormalities was preterm and had suffered multiple other CNS insults (83). They termed the condition “benign enlargement of the subarachnoid spaces.” Alvarez and colleagues reported that most children with this condition had normal development although more than half had a history of gross motor and language delay; however, all of the delays were transient and resolved by 15 to 18 months of age (05). Furthermore, they found a family history of macrocephaly in 88% of these infants, suggesting that benign familial macrocephaly and benign enlargement of the subarachnoid spaces were related. Reports of additional infants with benign enlargement of the subarachnoid spaces demonstrated that by the age of 2 years the excess fluid resolved, mild developmental delay was common but self-limited, family history of macrocephaly was common, and children often remained macrocephalic (124; 71; 38). A study showed overall good long-term outcome in these patients, with a slight increase in attention deficits with anecdotal histories of learning disabilities and behavior problems, but intelligence was within normal limits (93). Laubscher and colleagues reported similar findings, noting that 37 of 74 individuals with primary megalencephaly demonstrated variable degrees of developmental delay. The delay was transient in 18 of 37 patients (72). Halevy and colleagues reported that 9 out of 20 infants with macrocephaly due to benign enlargement of the subarachnoid space had mild hypotonia, which was persistent in 2 out of 10 (52).

Therefore, symptomatic familial megalencephaly may be only a transient finding, and during subsequent follow-up, such patients may exhibit no difference in the clinical course from clinically-insignificant or primary megalencephaly (43; 38). Many of the earlier reports showing increased rates of developmental delay and retardation associated with primary megalencephaly predate modern imaging modalities and laboratory testing and, thus, may include significant numbers of patients with secondary megalencephaly. Of note, a study demonstrated the low sensitivity and positive predictive value of head circumference monitoring in primary care clinics, suggesting that it is less head size and more other associated signs and symptoms that are important for determining which patients may have underlying pathology (27).

Thus, due to the availability of modern imaging and molecular testing, 2 more criteria are now considered essential for meeting the diagnostic criteria for asymptomatic (benign) familial anatomic megalencephaly (43):

(7) Radiographic evidence of normal or slightly enlarged ventricles in 1 of the family members with macrocephaly.

(8) Negative results of screening tests for metabolic disorders or lysosomal deficits.

Prognosis and complications

The outcome for megalencephaly is dependent on the underlying etiology. Cases due to neurocutaneous or genetic syndromes, lysosomal storage diseases, or inborn errors of metabolism are expected to follow the clinical course predicted in these specific entities (103). Unilateral or hemimegalencephaly has a relatively poor prognosis, with most patients demonstrating intractable seizures and global developmental delay (39; 40; 43). However, a case study in 2019 described a 55-year-old woman with normal cognition who presented with late-onset seizures secondary to hemimegalencephaly, possibly the oldest age for a first seizure in a patient with hemimegalencephaly (24).

In an older study of children referred for learning disabilities, 11% (compared to 4% of matched controls) were shown to have large head circumferences (121). A more extensive study of nonreferred children in resource programs in public schools revealed similar findings (122). A further study of children ages 6 to 15 years old showed that the presence of megalencephaly was associated with poorer performance on tasks of upper limb speed, visuomotor control, running speed, bilateral coordination, visuomotor integration, and naming fluency (110).

A large retrospective cohort using the Swedish Medical Birth Registry consisted of 1 cohort of 144,273 boys that contained 732 with primary megalencephaly and another cohort of 3,204 boys and girls with mental retardation. Primary megalencephaly was associated with low intelligence (odds ratio 1.31), and associated weight-normalized large head size at birth in both the mother and child was also found (99). Macrocephaly and megalencephaly have also been shown to be more common in children with autism (137; 10). Children with macrocephaly associated with developmental disabilities have also been reported to have increased risk for seizures (95).

In contrast, other studies have reported normal neurodevelopmental outcome in the majority of children with macrocephaly and megalencephaly (101; 26; 100; 75; 83; 05). Differences in these figures stem from variability in ages at evaluation, lengths of follow-up, exclusionary criteria, availability of neuroimaging or genetic testing, and how precise the definitions of macrocephaly and megalencephaly were.

Clinical vignette

A 6-month-old male was initially referred for evaluation for large head. His head circumference was in the 50th percentile at birth but had increased rapidly since birth to larger than 98th percentile, while his height and weight remained around the 50th percentile. His history was unremarkable except for mild gross motor delay (could only sit with support). He had no regression or loss of skills and no seizures and otherwise was healthy. His neurologic examination was normal except for mild hypotonia, with no noted stigmata of a genetic syndrome. Family history was notable for large heads on the paternal side, including his father. A head CT done before the visit demonstrated mild enlargement of the subarachnoid spaces and lateral ventricles. No further imaging or testing was ordered at that visit, and the patient was referred to physical therapy.

Over the next 6 months, the patient’s head size continued to increase, but the rate began to parallel the upper line of the curve. No new symptoms or signs developed, but the child continued to have mild gross motor delay. The father had a head CT done during the interval (after a minor fall), which showed no abnormalities but was consistent with megalencephaly. Based on the clinical history, examination, and family history, the child was diagnosed with benign enlargement of the subarachnoid spaces of infancy and benign familial anatomic megalencephaly.

Biological basis

Etiology and pathogenesis

Clinically significant megalencephaly is usually associated with an identifiable etiology or syndrome or has other associated anomalies. Broadly, megalencephaly may be due to anatomic or metabolic etiologies. In general, children with anatomic megalencephaly present at or close to birth with large heads, and after an initial further increase in head size, the rate of growth parallels the upper percentiles. This may be the presenting sign for genetic or neurocutaneous syndromes. In contrast, metabolic encephalopathy may not initially be associated with macrocephaly, and developmental and neurologic examination abnormalities may be more evident.

Metabolic megalencephaly may be due to leukodystrophies, most commonly Alexander disease, Canavan disease, and megalencephalic leukoencephalopathy with subcortical cysts. Furthermore, glutaric aciduria type 1, mucopolysaccharidoses, generalized gangliosidosis, and Tay-Sachs can also be causes of metabolic megalencephaly. The mechanism for both megalencephaly and brain dysfunction leading to neurologic abnormalities is edema and accumulation of metabolic products, abnormal in type or amount, without any inherent abnormality in the numbers or sizes of brain cells (43; 97; 138; 103).

Anatomic megalencephaly may be due to multiple different causes and is by far the more common type of megalencephaly. The most common is benign or familial or primary megalencephaly, described above. Anatomic megalencephaly may also be associated with genetic syndromes, which can be classified by associated anomalies. Those associated with somatic overgrowth include Sotos (cerebral gigantism), Beckwith-Wiedermann, fragile X, Simpson-Golabi-Behmel, and Weaver-Smith syndromes. Those associated with skeletal dysplasia include achondroplasia, Apert syndrome, and thanatophoric dysplasia syndrome. Those associated with neurocutaneous manifestations include neurofibromatosis type 1, nevoid basal cell carcinoma, macrocephaly cutis-marmorata telangiectasia congenita, Sturge-Weber syndrome, and PTEN hamartoma (including Bannayan-Riley-Ruvalcaba, Cowden, and Lhermimitte-Duclos) syndromes. Those associated with chromosomal abnormalities include Klinefelter and fragile X syndromes. Those associated with neuro-cardio-facio-cutaneous manifestations include Noonan, Costello, and cardiofaciocutaneous syndromes. Furthermore, FG, Greig cephalopolysyndactyly, and acrocallosal syndromes are additional causes. The genetic basis for most of the above described anatomic megalencephaly syndromes are known and include a wide range of gene mutations. These are characterized by increased neural or glial cell size, number, or both, leading to enlargement in the overall size of the brain. Enlarged numbers of cells can be due to increased proliferation or decreased apoptosis (43; 97; 138; 89).

Autism has also been shown to be associated with macrocephaly and megalencephaly. In fact, large head has been demonstrated to be the most consistent finding in children with autism. The abnormal rate of head growth appears to be more common during the infancy and early childhood age range (09; 102; 91). One report suggested that fetal head circumference may be a useful predictor of children at risk for developing autism (137). Additionally, it does not seem to matter which head growth chart is used, because they will be found to macrocephalic on all charts (90).

The pathogenesis of megalencephaly varies depending on the etiology. Megalencephaly in leukodystrophies and lysosomal storage diseases demonstrates the expected degenerative changes or the typical abnormal accumulations of byproducts of cellular metabolism. Anatomic megalencephaly secondary to various neurocutaneous syndromes may demonstrate neural and extraneural hamartomas as well as disturbances of cortical laminations with polymicrogyria or pachygyria observed grossly (Debaken and Priestly 1977; 30).

Discussion of the various anatomic and metabolic causes of megalencephaly is beyond the scope of this article, but a few will be addressed to provide insight into the current state of knowledge on pathogenesis and pathophysiology.

Alexander disease shows widespread hypertrophy of astrocytes with inclusions of dense protein aggregates (Rosenthal fibers) composed mainly of abnormal glial fibrillary acidic protein (GFAP), resulting in axonal loss and demyelination preferentially in the frontal lobes. The disorder results from autosomal dominant gain of function mutations in the GFAP gene. Gross examination of the brain demonstrates discolored and shrunken white matter despite the brain being heavier than normal. There is preferential involvement of white matter, causing macrocephaly, megalencephaly, seizures, spasticity, and early death. The abnormalities are noted only in astrocytes, and oligodendrocytes are uninvolved. Overexpression of GFAP in transgenic mice causes a fatal encephalopathy similar to the human condition (48; 97; 138; 123). A murine model has been created and will contribute further insight into the pathology of this disorder (125). In Canavan disease, the white matter demonstrates spongiform degeneration, with swollen and vacuolated astrocytes, apparently causing increased water spaces and decreased myelination especially in the arcuate fibers of the frontal and parietal lobes. The cause is deficiency of the aspartoacylase (ASPA) enzyme, which is important for N-acetylaspartic (NAA) hydrolysis to aspartate and acetate, resulting in elevated NAA levels in the urine in patients with Canavan disease (97; 138). APSA-deficient mice have been demonstrated to have abnormal oligodendrocyte maturation and myelination, although the mechanism is currently still under investigation (81).

Megalencephalic leukoencephalopathy with subcortical cysts (MLC) is autosomal recessive and most often due to mutations in the MCL1 gene (104). It presents with macrocephaly, diffuse mild cerebral white matter edema with sparing of central white matter and grey matter, and large subcortical cysts in the anterior temporal lobes (97; 138). Miles and colleagues published biopsy data of the frontal gyrus at age 15 months, showing normal gray matter (84). However, the subcortical white matter was noted to be pale due to the presence of prominent fine uniform vacuoles that resembled myelin, with a few interspersed myelinated axons and rare microglia. Blebs were noted forming from the outer and occasionally the inner layers of myelin surrounding intact axons, representing a possible source for the vacuoles. Fine uniform vacuolation of white matter with wide separation of myelinated axons was felt to be the hallmark of MLC in early childhood. An overlapping phenotype was described, where patients initially presented as MLC, but followed a milder course and lacked mutations in the MCL1 gene (van der Knapp et al 2010).

In most cases due to anatomic cause, megalencephaly results from the underlying neuropathology associated with the syndrome. Early autopsy cases of apparently idiopathic megalencephaly, however, demonstrated no obvious gross or microscopic pathology aside from an incidental cyst formation and surrounding gliosis in 1 case (06; 36). Biopsies of the left and right parietooccipital cortex of a patient with Soto syndrome of cerebral gigantism were also found to be normal (02).

Dekaban and Sakuragawa reviewed the literature of autopsy cases, in addition to 3 of their own, and found that of 31 patients with megalencephaly, 80% had documented large heads at less than 1 year of age, and over 60% had neurologic abnormalities. Histologic examination revealed no abnormalities in 8 patients. Abnormal cytoarchitecture (deficient cortical lamination, bizarrely shaped neurons, hypertrophy of neurons and glial elements, and microscopic heterotopias) was found in 11 patients. Grossly visible abnormalities (pachygyria, polymicrogyria, arrhinencephaly, etc.) were present in 10 brains (28).

Cases of hemimegalencephaly have provided more extensive data about the underlying pathogenesis, especially in autopsy cases where a less involved opposite cerebral hemisphere can be used as a control. Gross specimens at autopsy or following cortical hemiresection have demonstrated a firm, thickened cortex and white matter with polymicrogyria and an indistinct gray-white matter junction. Some cases have demonstrated lissencephaly and subcortical heterotopias (73; 128; 127; 79; 25; 31; 19; 39; 40). In some cases, polymicrogyria was present bilaterally (128; 25).

Microscopically, the brain in 1 case maintained normal cortical laminations but displayed thickening and the presence of scattered hypertrophied neurons and increased numbers of glial cells (128). Most cases demonstrated abnormal or no cortical laminations with large, abnormally oriented neurons and normal or coarse-appearing Nissl substance as well as uncharacteristic random arrangement of dendrites. Some had enlarged astrocytes and were binucleated, but mitotic figures were not seen (73; 14; 128; 127; 25; 31). Some samples had evidence of neurofibrillary tangles and neuronal cell types in uncharacteristically deeper areas (25; 31). On immunohistochemical analysis, 1 of 3 cases reported by DeRosa and colleagues was found to have staining of some cell groups against glial fibrillary acidic protein and vimentin, suggesting the cells were pluripotent (31).

Ronnett and colleagues, however, established a cortical neuronal cell line from a patient with unilateral megalencephaly in which the cells stained for neuronal markers but not for glial cell markers (106; 107). The cell line was positive for neuron-specific enolase and neurofilament but negative for glial fibrillary acid protein and S-100 protein. Furthermore, neurotransmitters normally found in abundance in the cerebral cortex were also found in this cell line (107). Tsuru and colleagues have immunohistochemically demonstrated abnormal expression of cell adhesion molecule L1 in 10 developing brains of children with hemimegalencephaly; they suggest that this is the cause of the cortical dysplasia and heterotopia (129). Flores-Sarnat and colleagues examined the cerebral tissue from 3 children after hemispherectomy for hemimegalencephaly, using histochemical and immunocytochemical markers of neuronal and glial maturation and identity. Histologic abnormalities of cellular growth and cytomorphology, such as "balloon cells," were present in both gray and white matter; disorganized tissue architecture was also observed. Many hypertrophic, atypical cells with enlarged processes of mixed or ambiguous lineage were noted, with immunoreactivity for both glial (GFAP; S-100beta) and neuronal proteins (microtubule-associated protein 2, neuronal nuclear antigen, chromogranin A, and neurofilament protein). Incomplete maturation was suggested by strong vimentin reactivity in neurons as well as glial cells and cells of mixed lineage. Electron microscopy demonstrated synaptophysin-reactive axons terminating on a minority of balloon cells and on most heterotopic single neurons in white matter, demonstrating that single heterotopic neurons are not synaptically "isolated," and, thus, may provide a focus for causing epilepsy. Oligodendrocytes appeared to be the least affected cells. The findings were felt to be similar to hamartomas of tuberous sclerosis. The authors concluded that hemimegalencephaly is a primary disorder of neuroepithelial lineage and cellular growth and that a migratory disturbance contributes to disorderly tissue architecture but is secondary. Of note, no pathologic differences were detected between isolated and syndromic forms of hemimegalencephaly (39; 40). Subsequent study of these cells may provide details of the mechanisms that result in the proposed unrestricted proliferation and partial migration arrest that lead to the abnormal cortical laminations and ectopic gray matter associated with unilateral megalencephaly.

Hemimegalencephaly and focal cortical dysplasia 2b have been suggested as a tauopathy in infants and children due to the upregulation of abnormally phosphorylated tau. This disrupts microtubules and can lead to a compensatory upregulation in neurons and glia. Sarnat and Flores-Sarnat hypothesized that hemimegalencephaly and focal cortical dysplasia 2b are the same disorder with different timings in the mitotic cycles of the neuroepithelium (112). They proposed that these disorders be classified as infantile tauopathies.

Cytomorphic data have demonstrated increased cell volume on the affected side compared to controls (31). Evaluation of nuclear and nucleolar volume in neuronal cells demonstrated increased volume of both, and biochemical analysis of tissue extracts showed increased DNA, RNA, and protein levels. This supports the idea that marked increase in transcription and translation is occurring on the affected side compared to the unaffected side (14; 79).

It seems clear that in the megalencephalic brain unaffected by abnormal storage, there has been a failure in control of cell numbers, size, and intercellular substance. Full understanding of the mechanisms involved in this control is still in the early stages. Growing evidence indicates that activation of N-methyl-D-aspartate receptors and other neurotransmitter-induced signals influences neuronal migration and cortical development (70). Studies of the nematode Caenorhabditis elegans have identified genes involved in the complex processes of programmed cell death, neuronal growth, and neurogenesis (35). During normal early brain development, a large number of cells are destroyed by a process called apoptosis and then engulfed by other cells. In Caenorhabditis elegans the number and type of cells that die and the removal of their debris is under genetic control of multiple different genes (65; 108). Trophic factors also affect cell growth and maintenance, and their expression is under the genetic control of genes such as c-fos and jun (55). Faulty genetic control of cell proliferation and differentiation might also lead to an abnormal accumulation of brain tissue (108). It should be emphasized that the processes involved in brain development do not act in isolation. For example, as a cell line develops there is inhibition of others through cell interactions mediated by neurogenic genes (54; 23; 108). In drosophila, a neurogenic gene called bigbrain (bib) is involved in transport across cell membranes, and faulty expression of this and other neurogenic genes leads to neural hypertrophy (54; 63; 108).

Several mouse models for megalencephaly have been created. In the insulin-like growth factor transgenic mouse, IGF-1 overexpression causes increased numbers of neurons and excessive myelin. The CD81-null mice also have enlarged brains due to increased numbers of astrocytes. Pten null mice have increased neural cell body size and large brains, and in humans this same tumor suppression mutation results in various cancers, hamartomas, and megalencephaly. The mceph/mceph mouse has postnatal progressive megalencephaly and seizures (34) and is due to potassium channel dysfunction (04). Pik3ca^H1047R and Pik3ca^E545K mouse lines with mutations at H1046R and E545K led to enlarged cell and nuclear size and increased proliferation. These mice showed significant brain overgrowth when compared to controls (109). Other genes involved in brain development include the paired box genes (PAX), which function as transcription factors. In mice, a point mutation at the Pax6 locus leads to malformations and increased volume of the forebrain germinal zone (113). A Pax6 rat has been used as a model for autism (133); in humans, mutations in this gene lead to eye and brain abnormalities (139).

Too many syndromes are associated with anatomic megalencephaly to list them here, but they are summarized in an excellent book chapter by AG Hunter (61). Many of these syndromes have been reported only in small numbers of patients, and delineation of the syndromes is ongoing. Updates have been published on some of them, such as the syndrome of megalencephaly, mega corpus callosum, and complete lack of motor development (58). Furthermore, new syndromes continue to be reported, such as macrocephaly, alopecia, cutis laxa and scoliosis (MACS syndrome), which is due to a mutation a RIN2, a protein important for endocytic trafficking (12), and mega-corpus callosum, polymicrogyria, and psychomotor retardation syndrome in consanguineous families (15). Mutations in the GTPase protein RAB39B have been reported as a cause of X-linked mental retardation associated with autism, epilepsy, and macrocephaly in 2 families (42). Additionally, some of the reported syndromes may actually represent variations of the same syndrome, such as megalencephaly polymicrogyria-polydactyly hydrocephalus (MPPH) and macrocephaly capillary malformation (MCM or MCAP) syndromes (49). MPP has been shown to be associated with chromosome 5 microdeletions in at least some patients (135), as well as mutations in PIK3CA (126; 66), PIK3R2, AK3T (88), MTOR (92; 85), and PTPN11 genes (33). These are all activating mutations in the phosphoatidylionositol-3-kinase/AKT/mTOR pathway (66). Studies have also shown de novo mutations in the downstream CCND2. This finding suggests that both upstream PI3K-AKT and downstream CCND2 mutational events cause the stabilization of cyclin D2, possibly contributing to megalencephaly (86). PTEN mutations have been suggested as a cause of autism and as mentioned above are associated with several megalencephaly syndromes (82). In fact, data suggests that children with autism and macrocephaly may have a subtype of autism more frequently associated with PTEN mutations (67; 80). Mutations in these genes have been shown to result in gain of function and overgrowth syndromes (105; 87; 60; 64). Alternatively, there have been cases with patients presenting megalencephaly capillary malformation symptoms without any identified genetic mutation, suggesting other underlying factors causing these syndromes (68).

Several authors cite aberrant control of proliferation, programmed cell death, migration, differentiation, and synaptogenesis of cells of the central nervous system as a “cause” for megalencephaly (25; 31; 01). This would account for the gross and microscopic pathology observed and allows for the suggestion that megalencephaly may be a phakomatosis (25). In addition, the onset of pathology can be postulated to occur at approximately 7 to 16 weeks’ gestation (11; 31). The immunohistochemical and biochemical data, however, suggest that more than a transient insult resulting in a static malformation has occurred. A possible defect in regulation of cell metabolism and the possibility of chromosomal DNA heteroploidy are postulated by some (14; 79). In any event, there is at least agreement that the pathophysiology is either multifactorial, a result of multiple superimposed events, or a singular insult with cascading consequences (14; 79; 31). Delineation of the possible genetic mechanisms responsible for the pathology observed in megalencephaly continues to be an area of active research (51; 62).

An entirely different mechanism may be responsible for primary or benign familial megalencephaly. Different pathophysiologic mechanisms have been proposed as causes of this form of megalencephaly. A number of authors have noted the autosomal dominant pattern of inheritance in familial megalencephaly (08; 26; DeMyer 1981; 75), although autosomal recessive (100; 96) and multi-factorial inheritance (07) have also been reported. Autosomal dominant modes of inheritance are also found in achondroplasia, neurofibromatosis, tuberous sclerosis, and many of the other forms of syndromic anatomic megalencephaly. The gangliosidoses and mucopolysaccharidoses are autosomal recessive except for Hunter syndrome, which is X-linked recessive (29; 45).

In addition to the discovery of the autosomal dominant inheritance pattern in most familial anatomic megalencephalies, many authors also recognize a mixture of idiopathic external hydrocephalus (more correctly called “benign enlargement of the subarachnoid spaces of infancy”) and benign familial megalencephaly within the same family and across several generations of macrocephalic individuals (08; 05; 72). Laubscher and colleagues proposed that “idiopathic external hydrocephalus” is actually a variant of familial megalencephaly, which subsequent studies have confirmed (see Extracerebral fluid collections in infants for summary and references).

The pathology of autism is currently an active area of investigation, and published data of brain biopsies of patients with autism showed evidence of multiregional dysregulation of neurogenesis, neuronal migration, and maturation, which the authors felt contributes to the heterogeneity of the autism clinical phenotype (136). Currently there are hypotheses that the zinc-metalloprotease-BDNF (ZMB) axis plays a role in the development of autism and contributes to megalencephaly, although this connection remains unclear (69).

Epidemiology

No clearly established incidence or prevalence has been reached despite of attempts to calculate these figures from available data. The selection of patients with occipital-frontal circumference 2 standard deviations above the mean leaves up to 2.5% of the general population who are "normal" with the diagnosis of macrocephaly. Furthermore, if imaging is not done, it is impossible to distinguish on clinical grounds alone whether the child has a large head due to a large brain (megalencephaly) or excessive fluid (including benign enlargement of subarachnoid spaces of infancy, another common and related condition causing macrocephaly in infants). These 2 conditions likely account for most of the familial and idiopathic forms of clinically-insignificant megalencephaly and macrocephaly (30). Prevalence rates reported from autopsy studies vary widely from 1 in 1146 to 1 in 50,000. Methods of measurement and handling of the brain prior to weighing may add to the variability (61).

Several authors have attempted to better define this population, and Lorber and Priestley found that of 557 children referred for evaluation of a large head, 109 had megalencephaly (20%). Five of these children were mentally retarded, leaving 95% of children with megalencephaly within the normal range of intellectual functioning. At least 50% of their cases had a familial history of large heads. A male-to-female preponderance of 4 to 1 was also demonstrated (75).

Other series have noted normal intelligence in at least 50% of patients with large heads whose growth curve parallels the normal pattern, with 10 of 15 children being male (100). In a series of normal children with large heads, positive family history was present in 73% of the cases, and males outnumbered females by 27 to 6 (26). Additionally, 50% of subjects with primary megalencephaly in a study by Laubscher and colleagues had normal intelligence and development, with an additional 24% overcoming transient developmental delays. In this series, males outnumbered females by 50 to 24, and a positive family history was obtained in 70% of the patients studied (72). In the study by Petersson and colleagues of 732 boys with primary megalencephaly, the estimated odds ratio for mental retardation was 1.33, and there was a correlation between maternal macrocephaly and a large head circumference of the child at birth (99).

As mentioned above, there are numerous neurocutaneous, overgrowth, and metabolic etiologies for megalencephaly. Autism has also been shown to be associated with megalencephaly and macrocephaly, with accelerated rates of head growth noted in early childhood (91; 137). Autism has a male predominance like primary megalencephaly and is also often diagnosed in the first few years of life. Certainly some of the children reported in earlier studies with mental retardation or developmental delays could have been autistic, given that many children with autism in prior decades were not diagnosed (117; 17).

Prevention

Genetic counseling and prenatal diagnosis are available for certain disorders where megalencephaly is an associated finding, though these procedures obviously do not prevent the disorder from having occurred in a child who has already been conceived. Benign familial megalencephaly has been demonstrated prenatally with the use of ultrasonography combined with family history, though again this does nothing to prevent the occurrence of megalencephaly in the child. The difficulty of distinguishing clinically-significant megalencephaly from insignificant megalencephaly, either before birth or during the early years of life, emphasizes the need to consider the latter when discussing prognosis of this detected finding (32). A study demonstrated that when fetal macrocephaly is associated with other brain or systemic abnormalities, syndromic macrocephaly can be diagnosed in utero (77). Other studies have demonstrated that isolated fetal macrocephaly is not a risk factor for long-term developmental problems (16; 03).

Differential diagnosis

Megalencephaly must first be differentiated from other causes of macrocephaly. Signs of increased intracranial pressure, rapidly increasing head size, or neuroimaging findings consistent with hydrocephalus, subdural hematoma, or thickened skull on bone windows or radiographs may direct the diagnosis away from megalencephaly and toward other causes of macrocephaly (30; 46). The differential diagnoses for macrocephaly are as follows:

• Hydrocephalus (noncommunicating vs. communicating) • Brain edema (toxic-metabolic) • Subdural fluid • Thickened skull • Megalencephaly

Once other causes of macrocephaly have been excluded, many possible etiologies of megalencephaly must be considered in the differential diagnosis, assisted by a comprehensive medical history and examination. Patients presenting with a normal head circumference at birth, slow acceleration of head size, and neurologic regression as well as abnormal neurologic findings, will be more likely to lead to a definitive diagnosis of metabolic megalencephaly. Other children may have a large head at birth with initial rapid acceleration of growth followed by a paralleling of the normal growth curve, may have delays in development but usually no regression, and are consistent with anatomic megalencephaly (30; 46; 43; 38).

Diagnostic workup

As with any clinical evaluation, a comprehensive history with a carefully directed physical examination will usually lead to the most appropriate diagnosis, or at least a more focused differential diagnosis. This clinical approach will often avoid an inappropriately extensive and expensive evaluation. The patient being evaluated for macrocephaly should first be confirmed by accurate occipital-frontal circumference measurements and plotting on standard graphs over multiple time points to meet criteria for enlarged head (94; 38). Rapid growth or crossing of percentiles suggests an underlying symptomatic etiology, such as hydrocephalus or clinically-significant megalencephaly. Appropriate neuroimaging should promptly be performed (94; 29; 30; 46; 138). There is evidence of early sudden death with those presenting with progressive cerebral enlargement, so early diagnosis is key in managing appropriate treatment (53).

Careful history (history of present illness, as well as past medical history and family and social histories) and thorough physical and neurologic examinations should be performed at the initial visit. In addition to the patient, examination of parents, siblings, or other apparently affected family members may need to be performed (including measurement of head size), looking for the stigmata of neurocutaneous syndromes or other systemic diseases. Based on thoughtful and thorough history and examination, more appropriate use of medical, neuroimaging, and neurogenetic tests may follow.

Some authors suggest that if there is no historical or clinical evidence of increased intracranial pressure in an otherwise apparently normal child with macrocephaly or megalencephaly and at least 1 parent has an statistically large occipital-frontal circumference, no further diagnostic workup may be necessary (26). Many, however, have considered neuroimaging with CT or MRI part of the routine workup of megalencephaly (57; 100; 30; 47). A published algorithm suggests initially separating children with large heads into 2 groups: evident syndrome versus no obvious genetic syndrome. Of the latter, those with no developmental delay would be diagnosed as familial macrocephaly or megalencephaly. If developmental delay was present in a child with no obvious syndrome, MRI or CT scanning then should be done (138). MRI certainly is helpful for identifying children with metabolic etiology for their megalencephaly (such as leukodystrophies or organic acidurias) and is able to reliably distinguish them from each other over time (138). Most children with anatomic megalencephaly have normal imaging, as expected (43).

MRI imaging will occasionally reveal dilated Virchow-Robin spaces, even in children with normal or relatively normal developmental and neurologic courses (56). The significance of these findings, as with apparently clinically-insignificant ventriculomegaly, should be interpreted carefully within the clinical context, often requiring close clinical sequential monitoring of the child’s neurodevelopmental course (50). One report demonstrated that once the sequelae of acute encephalopathy have been excluded, the diagnosis and prognosis of Virchow-Robin space dilatation likely represents a benign, non-progressive developmental disorder, although follow-up imaging may be warranted (22).

One third to one half of patients with benign familial macrocephaly may have mild to minimally dilated ventricles and slightly prominent subarachnoid spaces on neuroimaging, especially noted frontally. As mentioned above, this is referred to as “benign enlargement of the subarachnoid spaces of infancy.” It is commonly a transient finding, and genetically related to familial macrocephaly (see article for extracerebral fluid collections in infants). A report of intracranial intraoperative monitoring of 7 infants with benign enlargement of the subarachnoid spaces of infancy as well as ventriculomegaly demonstrated only 1 infant with mildly increased intracranial pressure, although the clinical significance of this was unclear (115).

Children with dysmorphic features or other physical findings suggesting a possible genetic etiology for their megalencephaly should have additional testing. Fragile-X PCR and high-resolution chromosomes as well as MRI imaging of the brain are warranted for any child with associated global developmental delay (118). Comparative genomic hybridization (CGH) may also be useful because it allows for the identification of very small duplications or deletions that may be casually related to the megalencephaly and developmental delay (119; 78). Continued advances in the resolution of CGH are allowing individuals to be diagnosed with increasing resolution, further demonstrating the complexity of evaluating data based on clinical and phenotypical diagnoses (18).

Management

Treatment in patients with megalencephaly is usually suggested by the underlying etiology. Every individual requires a thorough history and comprehensive physical and neurologic examination as well as appropriate follow-up. Examination of family members is useful in many cases for evidence of benign familial or primary megalencephaly, or for other clinical indications of a definable etiology. If no specific syndrome is identified and evaluation of the patient reveals a normal developmental history and neurologic exam, reassurance and observation, as well as neuroimaging in some cases, are adequate (43; 61; 138; 38). Patients with clinically-significant megalencephaly and associated developmental problems or seizures should receive appropriate imaging, referral for developmental interventions, genetic or metabolic testing, and/or anticonvulsant therapy (43; 138; 38). If these approaches do not control symptoms and seizures remain intractable, surgical resection of the focal megalencephaly should be considered. This should be done as early as possible to avoid suppressing normal childhood development (13).

Special considerations

Pregnancy

Cephalopelvic disproportion is a relatively common indication for cesarean section at delivery. However, there are no definitive data regarding the relationship between megalencephaly and cephalopelvic disproportion. Cesarean section for cephalopelvic disproportion had been performed in 3 of 15 children who were later evaluated for macrocephaly with head growth parallel to the normal curve (100). In another series, normal delivery was reported in only 44% of megalencephalic children diagnosed with primary megalencephaly. Specific labor and delivery difficulties were not given for all patients, but at least 1 was noted to require cesarean section for cephalopelvic disproportion (72).

The widespread availability of ultrasonography now provides an opportunity to assess in advance a potential for this problem. Benign familial megalencephaly has been diagnosed in utero with this technique. The authors wisely reiterate the importance of family history, physical exam of the relatives, and knowledge of the probable benign course in these cases when counseling families regarding detection of an enlarged head in utero (32; 16). Ultrasonography may be used to detect specific malformations or abnormal growth patterns suggesting syndromic macrocephaly, which can be helpful in counseling families (77). However, attempts to correlate fetal head or body size with risk of cesarean section have been disappointing (76).

Anesthesia

Complications with anesthesia have been documented in patients who are megalencephalic secondary to a mucopolysaccharidosis. The most common problem encountered with these patients is airway instability secondary to the extensive connective tissue involvement in these disease processes (59). Thoracic kyphoscoliosis and lumbar gibbus may make positioning difficult, but more concerning is the possible hypoplasia or absence of the odontoid process in Morquio and Hurler syndromes (120). Total cervical cord transection and compression of the medulla have been reported in connection with endotracheal intubation in these syndromes (20; 21). Other complications are related to additional characteristic pathologies associated with this group of diseases. Thus, more careful planning for and monitoring during anesthesia of children with in utero or early onset of macrocephaly or megalencephaly may be warranted.