ANOMALOUS DEVELOPMENT OF BRAIN STRUCTUREAND FUNCTION IN SPINA BIFIDA

MYELOMENINGOCELE

Jenifer Juranek1* and Michael S. Salman21Department of Pediatrics, Childrens Learning Institute, University of Texas Health Science Center at Houston, Houston, Texas2Section of Pediatric Neurology, Faculty of Medicine, Childrens Hospital, University of Manitoba, Winnipeg, Manitoba, Canada

Spina bifida myelomeningocele (SBM) is a specific type of neuraltube defect whereby the open neural tube at the level of the spinal cordalters brain development during early stages of gestation. Some struc-tural anomalies are virtually unique to individuals with SBM, including acomplex pattern of cerebellar dysplasia known as the Chiari II malforma-tion. Other structural anomalies are not necessarily unique to SBM,including altered development of the corpus callosum and posteriorfossa. Within SBM, tremendous heterogeneity is reflected in the degreeto which brain structures are atypical in qualitative appearance andquantitative measures of morphometry. Hallmark structural features ofSBM include overall reductions in posterior fossa and cerebellum size andvolume. Studies of the corpus callosum have shown complex patterns ofagenesis or hypoplasia along its rostral-caudal axis, with rostrum andsplenium regions particularly susceptible to agenesis. Studies of corticalregions have demonstrated complex patterns of thickening, thinning,and gyrification. Diffusion tensor imaging studies have reported compro-mised integrity of some specific white matter pathways. Given equallycomplex ocular motor, motor, and cognitive phenotypes consisting of rela-tive strengths and weaknesses that seem to align with altered structuraldevelopment, studies of SBM provide new insights to our current under-standing of brain structurefunction associations. ' 2010 Wiley-Liss, Inc.Dev Disabil Res Rev 2010;16:2330.

Key Words: Chiari II malformation; cerebellum; neuroimaging; structure;function; cortex; corpus callosum

Spina bifida myelomeningocele (SBM) is a special type ofneural tube defect whereby a failure in programmed fetaldevelopment (e.g., neural tube closure) at the level ofthe spine translates into substantially altered brain develop-ment, in both structural and functional domains. Both qualita-tive and quantitative studies have documented a remarkabledegree of heterogeneity among individuals with SBM in termsof size, shape, and appearance of the cerebellum, corpus cal-losum, and cerebral cortex. Similarly, a wide range of cogni-tive strengths and relative weaknesses among individuals withSBM are also documented in the published literature. As high-lighted below in the present paper, qualitative and quantitativeinvestigations of brain regions hypothesized to be associatedwith cognitive strengths and weaknesses in individuals withSBM provide new insights into current models of brain-behavior associations in general. Although hydrocephalus

commonly occurs in individuals with SBM, it will not be dis-cussed in detail here, and readers are referred to [Del Bigio,2010].

OVERVIEW OF STRUCTURAL FEATURESASSOCIATEDWITH SBM

Common structural characteristics associated with SBMinclude anomalous development of the skull, as well as infra-and supra-tentorial regions of the brain [McLone and Dias,2003]. Chiari II malformation (CII) is a complex congenitalanomaly that involves the midbrain, hindbrain (i.e., pons, me-dulla, and cerebellum) and cervical spinal cord [Harding andCopp, 2002; Barkovich, 2005], which occurs universally andexclusively in SBM [Wagner et al., 2002]. Prominent featuresof CII include a significantly smaller posterior fossa with itscontents crowded and distorted in appearance [Barkovich,2005]. Additional features of SBM can include tectal beaking,hydrocephalus, and corpus callosum hypoplasia and dysgenesis,but these features are variable. Although less studied, anenlarged massa intermedia and gyral malformations of theneocortex also occur commonly in SBM with CII.

The main features of CII are based on qualitative andquantitative descriptions of gross anomalies (Table 1), particu-larly midline brain structures [McLone and Dias, 2003]. Themesencephalic tectum is often distorted and stretched posteri-orly and inferiorly (see Fig. 1), which is frequently describedon magnetic resonance imaging (MRI) scans as beaking ofthe tectum. Tectal beaking is reported to occur in 75% of CIIcases, being particularly prominent in older patients with anextensive myelomeningocele [Fletcher et al., 2005]. Typicallyin CII, the brainstem is stretched down to the level of the fo-ramen magnum or cervical spine with narrowing in its anteri-

Grant sponsor: Eunice Kennedy Shriver National Institute of Child Health andHuman Development (NICHD) or the National Institutes of Health; Grantnumbers: P01-HD35946, R01 HD046609.*Correspondence to: Jenifer Juranek, Department of Pediatrics, University of TexasHealth Science Center at Houston, Childrens Learning Institute, 7000 Fannin Ste.2411, Houston, TX 77030. E-mail: jenifer.juranek@uth.tmc.eduReceived 10 January 2010; Accepted 1 March 2010Published online in Wiley InterScience (www.interscience.wiley.com).DOI: 10.1002/ddrr.88

DEVELOPMENTAL DISABILITIESRESEARCH REVIEWS 16: 23 30 (2010)

' 2010Wiley -Liss, Inc.

orposterior diameter. A cervico-med-ullary kink may be seen at the junctionbetween the cervical spinal cord and themedulla oblongata [Barkovich, 2005],which may lead to symptoms of spinalcord compression [Oaks and Gaskill,1992]. Although the inferior vermis fre-quently herniates down to the upper cer-vical vertebrae (C), at C2 or C4 level,the superior part of the cerebellar vermisherniates upward (Fig. 1). As describedby Barkovich [2005] and shown in Fig-ure 2, the cerebellum wraps around thebrainstem. In some cases, low insertionof the tentorium cerebelli, cerebellardysplasia, agenesis, or hypoplasia of thelateral cerebellar hemispheres also occurs[Gilbert et al., 1986; Harding and Copp,2002; Barkovich, 2005].

Morphology of the corpus cal-losum (CC) is also compromised duringearly embryogenesis in SBM [Barko-vich, 2005] as the CC programmed pat-tern of development usually occurs inutero between 7 and 20 weeks of gesta-tion [Barkovich and Norman, 1988].Complex patterns of CC dysmorphol-ogy occur along its rostro-caudal axis inSBM [Hannay et al., 2009]. Althoughpartial agenesis or dysgenesis and hypo-plasia (e.g., underdevelopment) are de-scriptive terms frequently used to com-municate type of dysmorphology, thegenu, body, splenium, and rostrum aresubdivisions of the CC used to commu-nicate where the dysmorphology hasoccurred [Fletcher et al., 2005; Milleret al., 2008]. An association has beenreported between spatial location of CC

agenesis and level of the vertebral col-umn where the myelomeningoceleoccurs (e.g., thoracic versus lumbar/sac-ral), with thoracic lesions being twice aslikely to have agenesis of the spleniumas lumbar/sacral level lesions [Hannayet al., 2009]. Thin regions of the CCare generally thought to develop sec-ondary to increased intracranial pressureand hydrocephalus, which develop in8090% of children born with SBM[Reigel and Rotenstein, 1994].

Theories and Mechanisms of ChiariType II Malformation

Over the years, several theorieshave been proposed to explain CII.Although no single theory fullyaccounts for all features of CII [Steven-son, 2004; Tubbs et al., 2008], the cur-rently most widely accepted theory isthe unified theory [McLone andKnepper, 1989]. According to thistheory, cerebrospinal fluid (CSF) leaksthrough the spinal defect into the intra-uterine environment. Failure of neuraltube closure hinders normal distensionof the embryonic ventricular systemthrough the pressure generated fromCSF buildup during the temporary pe-riod of spinal neurocele occlusion [Des-mond, 1982]. Lack of ventricular dis-tension in utero limits the normalgrowth of the bony elements of theposterior fossa, which significantlyreduces posterior fossa size [McLoneand Knepper, 1989; McLone and Dias,2003]. The contents of the posteriorfossa, which proliferate rapidly later[Griffiths et al., 2004], are constrainedby a smaller than normal posterior fossa,thereby lending the cerebellum tocompression and subsequent volumetricreduction as well as displacement and

herniation along its vertical and anteri-orposterior axes.

In light of the unified theory, severaltestable hypotheses to prevent CII fromdeveloping have emerged. The basic ra-tionale is that preventing CII will also pre-vent downstream consequences of CII[Walsh et al., 2001]. Specifically, one mightpredict that fetal surgical techniquesdesigned to repair the spinal defect in uterowould stop CSF leakage through the me-ningomyelocele, thereby preserving nor-mal distension of the ventricular system,and thus facilitating normal bone growth[see Bowman and McLone, 2010]. Re-ports of reduced or prenatal resolution ofhindbrain herniation following surgicalrepair of spinal meningomyelocele in bothanimal model systems [Meuli et al., 1995]and human patients [Adzick et al., 1998;Bruner et al., 1999; Sutton et al., 1999;Tulipan et al., 1999] are consistent withthe unified theory. Additionally, infantswith SBM who had fetal surgery to closetheir spinal defect have decreased symp-toms and signs of brainstem compression[Danzer et al., 2008]. Currently, an NIH-funded multicenter study is conducting aprospective randomized clinical trial (e.g.,Management of MyelomeningoceleStudy) to compare efficacy of prenatal fetalsurgery with standard postnatal repair onimproving neurological outcome andreducing anomalous brain development inSBM [Mitchell et al., 2004; Sutton andAdzick, 2004; Bowman et al., 2009; Bow-man and McLone, 2010].

Additional theories proposed toexplain CII that have not been widelysupported include: hindbrain dysgenesistheory, which postulated that the hind-

Table 1. Structural ChangesWithin the Posterior Fossain Chiari II Malformation

Small posterior fossaSmall cerebellar volume (especiallythe cerebellar hemispheres)

Reduced cerebellar weightDysplastic looking cerebellumExpanded midsagittal vermis areaVermis (6tonsils) herniation belowforamen magnum

Upward herniation of the superior vermisWrapping of the cerebellar hemispheresaround the brainstem

Enlarged tentorium incisuraDysplastic tentorium cerebelli withabnormally low insertion

Inferior displacement of thelower brainstem

Tectal beakingMedullary-cervical kinkAqueduct stenosisSmall and inferiorly displaced fourthventricle

Enlarged foramen magnum

Fig. 1. T1-weighted midsagittal MRI in achild with Chiari type II malformationillustrating herniation of the inferior cere-bellar vermis (horizontal arrow) and tectalbeaking (vertical arrow). The fourth ven-tricle is small in size. Fig. 2. T1-weighted axial MRI at the

level of the pons in a child with ChiariType II malformation illustrating wrappingof the cerebellar hemispheres around thebrainstem (arrow).

24 Dev Disabil Res Rev Development of BRAIN Structure and Function JUranek And SALMAN

brain was underdeveloped in SBM; thetraction theory in which downwardtraction of the lower spinal cord wasthought to pull down the hindbraininto the spinal canal; the overgrowththeory, which suggested there is over-growth of the neural tissue prior to clo-sure of the neural tube; and the hydro-dynamic theory in which there isdownward herniation of the hindbraincaused by the raised intracranial pressurefrom the hydrocephalus [Gardner et al.,1975; Gilbert et al., 1986; McLone andKnepper, 1989; Harding and Copp,2002].

Recent investigations have triedto bridge both ends of the myelome-ningocele problem: the open neuraltube defect in the spinal cord andreduced posterior fossa size in the cra-nium. As suggested by Williams [2008],both genetic and environmental factorsare likely to have key roles in the etiol-ogy of posterior fossa hypoplasia in CII,which together with low-spinal pres-sure, contribute to the development ofCII [Williams, 2008]. In Sarnats molec-ular genetic explanation for CII abnor-malities, he suggested that abnormalrhombomere formation and ectopicexpression of homeobox genes of therostrocaudal axis of the neural tube mayexplain CII features [Sarnat, 2000].Clearly, additional research is needed toclarify the mechanism(s) mediating therobust association between the forma-tion of the myelomeningocele in thespine and the CII in the cranium.

As enthusiasm for genotypephe-notype studies focused on the cerebel-lum continues to spread, our knowledgeof molecular cascades underlying neuraltube closure and posterior fossa devel-opment will expand accordingly [seeAu et al., 2010]. Recent reports suggesta transcription factor expressed in thehead mesenchyme, FOXC1, is requiredfor normal skull development [Riceet al., 2003] and normal cerebellar de-velopment [Aldinger et al., 2009b]. Var-iations in the amino acid sequence ofFOXC1 can result in a spectrum ofphenotypes in cerebellar morphologyvia defective mesenchymal signaling[Aldinger et al., 2009a]. FOXC1 is areasonable candidate for further study asit is expressed near the time of neuraltube closure and is localized to the mes-enchyme where skull development,brainstem, and cerebellar developmentcould all be influenced by variability inits expression levels [Aldinger et al.,2009b]. However, definitive mecha-nisms underlying CII development stillremain to be elucidated.

The Cerebellum: QuantitativeStudies

Studies of the cerebellum haveconsistently reported reduced cerebellarsize in SBM, which is even evidentearly in fetal development. In a 16-week-old fetus with SBM, the neuraltissue in the posterior fossa was foundto be half the size of the neural tissue inthe posterior fossa in a control fetus[Brocklehurst, 1969]. In young neo-nates, significant reduction in cerebellarweight was found in 100 children withSBM, who died within 4 weeks ofbirth, in comparison with cerebellarweight of 60 children with normal ver-mis who died from other causes [Var-iend and Emery, 1973]. Subsequentdegeneration of the cerebellum in CII,also referred to as vanishing cerebel-lum, in CII has also been reported(e.g., in Seners [1995] study of fourchildren, age 1 month to 4.5 years,using magnetic resonance imaging(MRI) of the brain). Microscopic stud-ies have described reduced cell numbersand DNA content, particularly in theinternal granular layer of the inferiorvermis, in 100 children with SBMcompared with 120 controls [Emeryand Gadsdon, 1975]. More recently, mi-croscopic analyses of herniated cerebel-lar tissue have documented granule andPurkinje cell loss, reduction and gliosisof the folia, and myelin depletion [Har-ding and Copp, 2002].

Distance measurements, planime-try (e.g., area measurements), and bio-metric (e.g., shape) analyses of the pos-terior fossa and its contents have beenperformed in MRI studies of individu-als with SBM to quantify abnormalities,document variability, and predict func-tion or prognosis. In SBM, the longesttransverse and longitudinal distancesacross the midsagittal vermis indicatedexpansion, not reduction, relative toage-matched controls [Salman et al.,2006a]. In the same study, mean areameasurements of the midsagittal vermis(anterior, posterior, and inferior subdi-visions) were significantly increased by25% in SBM compared to normal con-trols despite a 9% reduction in posteriorfossa area measurements in SBM[Salman et al., 2006a]. Using shapeanalyses, Tsai et al. [2002] quantitativelydocumented the tremendous variabilityacross individuals with SBM in thedegree to which the hindbrain is mal-formed. According to their findings,shape changes in bone and vermiandescent co-occur in a relational fashion.However, herniation of the cervico-medullary junction appears to be inde-

pendent of vermian herniation, therebysuggesting different etiologies for her-niation of the vermis and herniation ofthe cervico-medullary junction.

Early volumetric studies of thecerebellum in a group of individualswith SBM used noninvasive MRImethods to quantitatively assess groupdifferences between 52 children withSBM and 16 age-matched controls[Dennis et al., 2004]. Although thisstudy found significantly reduced cere-bellar volumes in the SBM group, italso described a significant effect oflesion level on cerebellar gray matter(GM) and white matter (WM) volumes.Consistent with neurological outcomes,where individuals with lower level spi-nal lesions are less impaired than thosewith thoracic level lesions, higher spinallesion levels exhibited greater reductionsin cerebellar volumes of GM and WM(37%) than those with lower levellesions (22%) [Dennis et al., 2004].Another study of children with SBM[Fletcher et al., 2005] reported reducedcerebrum and cerebellar volumes inchildren with upper (n 5 25) in com-parison with lower (n 5 61) spinallesion levels. Together, these studiesdemonstrate that upper level spinallesions in SBM are a marker of moresevere brain dysmorphology and areassociated with poorer cognitive andbehavioral outcomes [Dennis et al.,2004; Fletcher et al., 2005].

A recent volumetric MRI studyused a simple approach for parcellatingthe cerebellum into meaningful subu-nits: the corpus medullare, the anteriorlobe, and the posteriorsuperior andposteriorinferior subdivisions [Juraneket al., in press]. Despite the overallreduction in cerebellar volume in theSBM group, absolute and relative vol-ume of the anterior lobe was signifi-cantly increased relative to a typicallydeveloping comparison group. Althoughthe absolute volume of the corpusmedullare was significantly reduced inthe SBM group, the relative volume ofthe corpus medullare was not signifi-cantly different from the typically devel-oping group when corrected for totalcerebellar volume. As shown in Figure3, the smaller cerebellum observed inindividuals with SBM is not simply, lin-early reduced equally across all regions,but rather disproportionately enlarged inthe anterior lobe, reduced in the poste-riorinferior subdivision, and not signif-icantly different in the corpus medullareor the posteriorsuperior subdivision.

The functional significance ofregionally specific patterns of volumet-

Dev Disabil Res Rev Development of BRAIN Structure and Function JUranek And SALMAN 25

ric enlargement and reduction in cere-bellar subdivisions remains to be estab-lished. Although the anterior lobe isgenerally associated with motor func-tion, it is not yet clear whether a largeranterior lobe confers advantages or dis-advantages in the motor function do-main. Individuals with SBM demon-strate a wide range of deficits in motorskills requiring strength or speed[Hetherington and Dennis, 1999] ordynamic regulatory control [Sandleret al., 1993; Jewell et al., 2009]. In con-trast, they perform pretty well on motorskills requiring learning and adaptation[Colvin et al., 2003; Edelstein et al.,2004; Dennis et al., 2006a]. Thus,ongoing studies investigating these typesof brain structurefunction associationsare ideally suited to provide newknowledge and insight for continuedprogress in hypothesis-driven research.

Consequences of Chiari Type IIMalformation on Hindbrain Motorand Visceral Functions

Intrinsic brainstem abnormalitiesand pressure on the brainstem from acrowded posterior fossa and hydroceph-alus give rise to a diverse range ofsymptoms such as headaches, apnea,

bradycardia, dysphagia, torticollis, spas-ticity, and impaired auditory evokedpotentials [Wagner et al., 2002; Henri-ques Filho and Pratesi, 2006; Bowmanand McLone, 2010]. Brainstem com-pression occurs in a third of patientswith CII and may be fatal in about athird of these patients [Stevenson,2004]. Surgical decompression of theposterior fossa for symptomatic CII isrequired in 817% of subjects withSBM [Wagner et al., 2002] and usuallyleads to clinical improvement. Maternalfetal surgery for SBM may also reducethe incidence and severity of brainstemdysfunction [Danzer et al., 2008].

Several eye movement abnormal-ities occur in CII including impairedsmooth ocular pursuit, saccadic (i.e., fasteye movements) dysmetria, impairedperformance of the vestibular-ocularreflex, various forms of pathologicalnystagmus (abnormal ocular oscilla-tions), strabismus, and internuclear oph-thalmoplegia (a type of gaze palsy thatcauses double vision) [Leigh and Zee,2006]. Recording different classes ofeye movements in CII has shed newlight on motor function of the hind-brain in CII in relation to the anatomi-cal changes in this malformation. Mid-

sagittal vermis expansion, including thesignificant expansion of vermis lobulesVI-VII area and preserved medial cere-bellar volume corresponded to sparingof eye movements processed by the ver-mis, despite the overall reduction incerebellar size that accompanies the de-formity of CII [Salman et al., 2009].These eye movement systems includesaccadic accuracy, saccadic adaptation (aform of ocular motor learning), andsmooth ocular pursuit [Salman et al.,2005, 2006b, 2007] In contrast, CIIpatients with abnormal ocular motorfunctions did not have a significantlyexpanded midsagittal vermis area. Fur-thermore, their midsagittal vermis areaoccupied a smaller midsagittal posteriorfossa area, and their cerebellar volumeswere smaller than in patients with CII,who had normal eye movements[Salman et al., 2009].

Therefore, the overall reductionin posterior fossa and cerebellar sizes inCII does not affect cerebellar structuresor functions uniformly, because there isevidence for functional ocular motorcircuits within the deformed hindbrainstructures in many patients with CII. Inaddition, saccadic adaptation stilloccurred in CII despite the small anddysplastic cerebellum and was evidenteven in patients with eye movementabnormalities [Salman et al., 2006b,2009]. Other types of motor learninghave also been reported to be preservedin patients with CII including adapta-tion on reaching or ballistic movements,mirror drawing performance, andweight judgments [Colvin et al., 2003;Edelstein et al., 2004; Dennis et al.,2006a].

The Corpus CallosumEven before sophisticated imaging

methods became available for quantita-tive analyses of WM in SBM, aberrantdevelopment of the corpus callosum in7090% of SBM cases [Barkovich,2005] made it a key focus of early qual-itative imaging investigations. Dysgene-sis of the CC is not a unique feature toSBM, as it also occurs in other neuro-genetic syndromes [Ettlinger, 1977].Yet, along with hydrocephalus, CC dys-morphology was one of the earliest visi-ble signs that incomplete neural tubeclosure at the level of the spinal cordhad serious consequences extendingbeyond the tentorium.

As reported in qualitative studiesof the CC, morphological variabilityis remarkable across individuals withSBM [Reigel and Rotenstein, 1994].Although some of this variability may

Fig. 3. The cerebellum has been parcellated into three primarily gray matter compartmentsas shown in sagittal (left panels) and coronal (right panels) views. Boundary demarcationsbetween compartments are indicated by bright white lines (anterior lobe, superior-posterior,and inferior-posterior subdivisions) in three representative subjects (A, B). A: Typically develop-ing child with cerebellar subdivisions numbered as: 1, anterior lobe; 2, superiorposterior; and3, inferiorposterior. B: Child with SBM. The central corpus medullare was also parceled sepa-rate from the other three parcels for quantitative analyses. The smaller cerebellum observed inindividuals with SBM is not simply linearly reduced equally across all regions, but rather dispro-portionately enlarged in the anterior lobe, reduced in the posteriorinferior subdivision, andnot significantly different in the corpus medullare or the posteriorsuperior subdivision.

26 Dev Disabil Res Rev Development of BRAIN Structure and Function JUranek And SALMAN

be related to lesion level, lesion leveldoes not account for all of the hetero-geneity [Fletcher et al., 2005]. Recently,26 regional patterns of CC morphologyhave been carefully documented in asample of 193 children with SBM basedon qualitative assessment of CC appear-ance across four subdivisions: rostrum,genu, body, and splenium [Hannayet al., 2009]. Whereas more than 50%of the study sample exhibited partialdysgenesis in either the rostrum (21%)or splenium (10%) or both regions(22%), only 4% of the study samplehad normal-appearing CC morphologyacross all four subdivisions. The remain-der of the study sample did not exhibitany dysgenesis, but rather displayed var-ious regional patterns of hypoplastic andnormal appearing subdivisions of theCC.

In early quantitative MRI investi-gations of the corpus callosum in SBM,cross sectional area measurements wereobtained from a single mid-sagittal slicein each participant [Fletcher et al.,1992, 1996; Hommet et al., 2002;Kawamura et al., 2002]. On average,the area measurement of the CC was

small in children with SBM comparedwith controls. In all these studies, adiverse spectrum of CC morphologyacross SBM cases was reported; a mi-nority of cases appearing to have nor-mal CC size and morphology, and themajority of cases appearing abnormalwith either hypoplastic features and/orpartial agenesis. Consistent with a ros-tral-caudal developmental course of theCC [Barkovich and Norman, 1988],the genu was reported to be relativelywell preserved in individuals with SBM,with hypoplastic features occurring inthe body, and agenesis most prominentin the isthmus/splenium subregion[Kawamura et al., 2002].

Given the high incidence of pos-terior CC dysmorphology in individualswith SBM, it is not too surprising thatsignificant deficits commonly occur inperformance of cognitive tasks requiringrecruitment of posterior networks,interhemispheric transfer, and transfer oflanguage information [Huber-Okrainecet al., 2005; Dennis et al., 2006b; Han-nay et al., 2008]. Thus, processing ofsemantic information is performed rela-tively easily, presumably due to a well-

preserved genu as suggested by resultsof split-brain surgical patient studies[Sidtis et al., 1981]. However, idiomcomprehension is negatively impactedby hypoplastic CC morphology[Huber-Okrainec et al., 2005]. Eveninterhemispheric processing of simpleauditory signals, as used in a dichoticlistening task, is influenced by spleniumdysgenesis, but not CC hypoplasia[Hannay et al., 2008].

The CortexOriginally described nearly 20

years ago using pneumoencephalogra-phy and cerebral tomography imaging,posterior thinning of the cortical mantlein SBM was an important discovery thatseemed to explain behavioral deficits inposterior network function [Denniset al., 1981]. Using en bloc regionaldefinitions in their volumetric analysesof structural MRIs, Fletcher et al.[2005] described significant reductionsin both GM and WM volumes poste-rior to the genu of the corpus callosumin SBM. More recently, high-resolutionsurface-based cortical reconstructionanalysis methods developed by Fischland Dale [2000] have been used. Thisstudy reported significantly reducedcortical thickness in posterior regionsand significantly increased corticalthickness in frontal regions based ongroup comparisons between an SBMgroup and healthy controls matched onage and gender [Juranek et al., 2008].These findings of altered spatial patternsof cortical thickening and thinning sug-gest a potential structural correlates ofcognitive strengths and weaknesses gen-erally exhibited by individuals withSBM.

Additional morphometry measuresderived from surface-based analyses of theneocortex in individuals with SBM haveincluded cortical GM volume and corticalcomplexity (e.g., gyrification). Using a3D local gyrification index (LGI) measureof cortical complexity developed bySchaer et al. [2008], spatial patterns ofcortical gyrification have been contrastedwith patterns of cortical thicknessbetween a group of individuals with SBM(n 5 74) and a comparison group of typi-cally developing study participants (n 531). As shown in Figures 4 and 5, inferiorparietal and posterior temporal areasexhibiting cortical thinning are also char-acterized by increased cortical complexity.In frontal regions, SBM exhibits increasedcortical thickness bilaterally in the follow-ing regions: inferior frontal, middle fron-tal, medial orbitofrontal, rostral anterior

Fig. 4. Significant group differences in cortical thickness between individuals with SBM (n 574) and a comparison group of typically developing (TD) study participants (n 5 31). Hotcolors indicate cortical thickness is greater in the SBM group relative to the TD group; coldcolors indicate cortical thickness is reduced in the SBM group relative to the TD group. Colorscale bar indicates significance values (corrected for multiple comparisons) ranging from P