Neurobiology of developmental dyslexia: Part 1: A review of evidence from autopsy and structural neuro-imaging studies

Developmental dyslexia (DD) is a languagebased neurological disorder which impairs reading ability but does not result from low intelligence, lack of motivation, sensory impairment, or inadequate instruction. Although the neurological basis of dyslexia has long been assumed, the exact nature of the altered brain structure associated with DD remains unknown and has been a subject of autopsy and neuro-imaging research. Autopsy studies provide consistent evidence of symmetry of the planum temporale (PT), thalamus and cortical malformations, whereas results from structural imaging studies such as computed tomography (CT) and magnetic resonance imaging (MRI) are inconsistent. To address the possible etiology of DD, this paper reviews evidence from autopsy and structural imaging studies on developmental dyslexia and discusses possible methodological sources of some inconsistent results. The role of the optometrist in the multidisciplinary management of dyslexia is highlighted. (S Afr Optom 2011 70(4) 191-202)


Introduction
In 1968, the World Federation of Neurologists defined dyslexia as "a specific learning disability that is neurological in origin.It is characterized by difficulties with accurate and/or fluent word recognition and by poor spelling and decoding abilities.These difficulties typically result from a deficit in the phonological component of language that is often unexpected in relation to other cognitive abilities and the provision of effective classroom instruction.Secondary consequences may include problems in reading comprehension and reduced reading experience that can impede the growth of vocabulary and background knowledge" 1 .
From a historical perspective, converging reports on the neurobiology of developmental dyslexia reveals that earlier understandings of the neurobiology of development dyslexia was derived from studies by Morgan, Hinshelwood, Dejerine Wernicke and Broca 2, 3 .Karl Wernicke, in 1874 (cited by Nakada) 3 , reported that a lesion in the left hemisphere resulted in a unique language disorder characterized by comprehension difficulties (Wernicke's aphasia).The lesions associated with Wernicke's aphasia are typically found on the superior surface of the temporal lobe between the auditory cortex and angular gyrus-the area of the cortex later known as the Wernicke's area 2, 3 , the same area where a language-relevant structure, the planum temporale (PT) is located 4 .According to Habib 2 , it was first reported by Dejerine in 1891 that damage to the same region of the brain (angular gyrus) resulted BSc(Hons)Optom OD (Benin) MOptom (UKZN) PG Cert Mod L/Vision (London) PG Cert Mod OcDiseases Ant/Seg (London) 192 in variable degree of impairments in reading and writing, which suggested that the left angular gyrus plays a role in reading impairments.Earliest descriptions of dyslexia therefore have related the disorder mainly to cerebral pathology.Dyslexia is a language-based disorder 5,6 of neurobiological origin 7,8 although the exact brain structure involved in the etiology of DD is still poorly understood.Consequently, studies investigating the neurobiological basis of dyslexia have focused on those neural systems subserving language, which are located in the Perisylvian association cortex in the left hemisphere of the brain 9 ,10 .An approach to determining the neuroanatomical abnormality in a behavioural condition such as dyslexia is to study related microscopic (neuronal level) or macroscopic (molecular level) as structural abnormalities might correspond to the behavioural abnormalities characteristic of the disorder 11 .An understanding of the neurobiological basis of dyslexia will guide diagnosis and intervention 7 .Certainly, dyslexia is primarily not caused by vision anomalies but optometry plays an important role in addressing the vision needs of a dyslexic child, which then makes intervention easier.Consequently, despite the role of optometry in the multidisciplinary management of dyslexia, information on dyslexia is sparse in optometric literature.This paper therefore, serves to contribute to the understanding of dyslexia.
Using the available empirical evidence, studies of brain morphometry (measurement of brain structures) derived from postmortem and structural imaging studies are reviewed with the aim to address the question of whether anatomical deviations in patterns of brain asymmetry (a normal brain is asymmetrical) characterizes the brains of dyslexic persons.The focus of this review is on the PT (at the macroscopic level), cortical malformations (at the microscopic level) and the thalamus.The review is presented in three sections.The first section will review evidence from autopsy studies on the PT, cortical malformations and the thalamus and the second section will review studies using structural imaging studies on the PT.In the final section, the methodological limitations that contribute to inconsistent results are discussed.First, it is necessary to present a brief review of basic neuro-anatomy of language-relevant areas of the brain, which will enhance an understanding of subsequent papers on the neurobiology of DD.

Basic neuro-anatomy of language-relevant structures in the brain
The brain, consisting of billions of neurons (nerve cells), which communicate with each other along an electrochemical path 5, 12-13 is divided into left and right hemispheres, connected by a bundle of nerve fibers, the corpus callosum.The cerebrum is the largest part of the brain and has a typical pattern of cellular arrangement (Cytoarchitecture) 5,[12][13] .The part of the brain that is visible on the surface is called the cerebral cortex 5 and is the layer of the brain often referred to as gray matter.The cortex covers the outer portion of the cerebrum and cerebellum 5 .The regions marked around the Sylvian fissure (which separates the temporal lobe from the frontal and parietal lobe) are called the perisylvian region 5,[12][13] .The perisylvian region is on the surface of the brain (the cortex), and holds the majority of language tissue.The lower bank of the Sylvian fissure contains the primary auditory cortex on Heschl's gyrus and auditory association cortex on the planum 5, 12-13 .Heschl's gyrus lies on the anterior boundary to the PT.The cerebral cortex has folds that allow the cortex to fit compactly into the skull.The folds of the cerebral cortex give the surface of the human brain its wrinkled appearance ("ridges and valleys") 5,[12][13] .The ridges are called gyri and the valleys, sulci or fissures (small and large respectively).The largest sulci are sometimes called fissures.Several deep sulci divide the cortex into four areas: frontal, parietal, occipital, and temporal lobes 5,[12][13] .The temporal lobe is a region of the cerebral cortex that is located beneath the Sylvian fissure on both cerebral hemispheres of the mammalian brain 5,[12][13] .The auditory cortex, located in the temporal lobe is a highly organized processing unit of sound in the brain.The cortex area is the neural base of language in humans 5, 12-13 .Another important anatomical structure is the angular gyrus-which is a part of the brain in the parietal lobe, that lies near the superior edge of the temporal lobe and is involved in a number of processes related to language, and cognition 5,[12][13] .Language-related information comes from the auditory cortex (for spoken language) or from the visual cortex (for written language) 3 .Also, relevant to the discussions of the neurobiology of dyslexia are the concept of lateralization, cerebral dominance and asymmetry.Brain lateralization is the phenomenon whereby a given function is preferentially controlled by one side of the brain relative to the other that is, two halves of the brain serve different functions.Cerebral dominance is an anatomical term that explains the superiority of one side of the brain for a particular function.Each half of the brain is dominant for several functions, for example, the left side is usually dominant for language, the right side for certain musical and spatial abilities.Brain asymmetry refers to a proportional difference between the right and left hemispheres of the brain 9 .As the PT is central to the subject of this review, a brief description of the PT is therefore warranted.

The planum temporale (PT)
Although there is controversy on how to define the exact anatomical borders, the PT has been conventionally described as a part of the temporal lobe, known to be relevant in language functions 4,[14][15][16] .The PT is a roughly triangular structure, which lies on the superior surface of the temporal lobe (just adjacent to the first Heschl's gyrus) inside the Sylvian fissure (SF) and it is a region of the cortex that falls within the Wernicke's area, on the left hemisphere of the brain [14][15][16] .
According to Kusych et al 17 , although a relationship between the PT and language function was first inferred by Wernicke, research interest in the studies of brain morphometry of the PT were stimulated by the landmark study of Geschwind and Levitsky 4 .Gescwhind and Levisky 4 performed autopsy examinations of the length of the PT in 100 normal brains and found it to be longer on the left in 65%, symmetrical in 25%, and shorter on the left in only 10% of the sample.The authors therefore hypothesized that the larger left PT might be an anatomical indicator for the specialization of the left hemisphere for language 4 .Since the PT lies within the classical posterior language region in the left hemisphere, defective development of language-related abilities in dyslexic subjects was thought to be related to a lack of asymmetry [18][19][20] .Individuals with unusual asymmetry (symmetrical) of the PT are at risk of being dyslexic because their left hemispheres are not as structurally adapted for language as they are in individuals with usual patterns of asymmetry 20 .The PT is also considered an anatomical structure in language because it serves as the foundation for the auditory association cortex (a posterior language area in the left hemisphere) and it is part of the classical Wernicke's area 14,20 .Also, the PT appears to have cytoarchitectonic properties which are necessary to relate the anatomical structure to functional significance 20,21 .Furthermore, since the PT asymmetry is apparent by the 29th to 31st week of gestation, abnormalities in this brain region may suggest a disruption of neurodevelopmental processes involved in establishing functional hemispheric lateralization 16 .Lastly, neuroimaging studies 3,17,22 have shown that the PT plays an important role in language processing.Given the role of the planum temporale in language function and dyslexia being a languagebased disorder, histo-pathological and neuro-imaging studies have been focused on the PT.These studies are reviewed below.

A review of the literature
Studies of brain morphology tend to follow a chronological pattern.Studies conducted before the 1980s were mainly autopsy examinations, studies in the early 1990s were structural imaging studies (CTs and MRIs) while studies conducted from the late 1990s till date were mainly on functional imaging (conducted while an individual performs a specific task such as reading).
(1) Autopsy Studies Autopsy studies investigate the brains of developmental dyslexics that have died from illness or accident.Anatomically, dyslexic brains have been found to be structurally different from those of non-dyslexic brains [23][24][25][26][27][28][29] .Evidence from autopsy studies revealed anomalies at two levels: microscopic (neuronal level) differences on cortex that includes ectopias, dysplesia and microgyria and macroscopic differences which include symmetry in the PT [23][24] .
The study by Drake 30 in 1968 was the first autopsy study to reflect on the neuro-pathological basis of developmental dyslexia.Drake 30 reported the case of a 12-year-old child, who had marked difficulties in reading comprehension, problems in arithmetic, poor spelling and recurring frontal headaches.Postmortem (autopsy) examination of the brain revealed a massive hemorrhage in the inferior vermis of the cerebellum, The South African Optometrist ISSN 0378-9411 194 an abnormal convulated pattern bilaterally in the parietal lobes, a thinned corpus callosum, and ectopic neurons deep in the white matter.There was no report on the asymmetry of the PT.
(i) Findings on the PT Galaburda and colleagues [23][24][25][26][27][28][29] have been the main researchers in the area of autopsy studies of persons with documented histories of dyslexic brains and have studied several dyslexic brains until date.In 1979, Galaburda and Kemper 21 reported the postmortem results of case of a 20 year old dyslexic man who had a family history of dyslexia.In 1985, Galaburda et al 23 reported three additional postmortem findings of three males, aged 14, 20 and 32 years respectively.Later in 1989, Humphrey et al 24 (working with Galaburda' et al) reported the neuro-patholological findings in three female dyslexic patients.So far, a total of 11 autopsied dyslexic brains (eight male, three females) have been studied by Galaburda and colleagues [23][24][25][26][27][28][29] .A remarkable finding with postmortem examination by Galaburda and colleagues is that all eight postmortem cases that examined the PT (specifically) of dyslexic subjects (six males and two females) had symmetrical PT 25 (normal brains have asymmetrical PT) 4 .Galaburda argued that because less than 33% of ordinary brains have symmetrical PT, the probability of encountering eight dyslexic brains with symmetrical PT by chance is minimal 25 .The reported symmetry in the dyslexic subjects resulted from a bilaterally large PT, (compared to those in normal brains) rather than from a reduction on the left PT 31 .In addition, this enlargement of the PT suggested anomalous brain development during the later stages of (prenatal) corticogenesis, which may result in abnormally high levels of surviving neurons with a subsequent restructuring of cortical architecture 23 .This, in effect, means that the mechanism for eliminating unwanted cells is defective in dyslexia forming a kind of "miswiring" in the brain 23 .

(ii)
Cortical malformations In addition to the findings of symmetrical PT, Galburda et al 21,[23][24] also observed that the dyslexics' cerebral cortexs contained some cortical malformations in the form of neuronal ectopias, architectonic dysplesias (focally distorted cortical architecture) 32 and microgyria (abnormal infoldings) 21,23,33 in eight of their male and one of the female dyslexic postmortem samples 33 .Ectopias are misplaced cells comprising brain cells and glia (supporting cells) that are located in areas of the cortex that should not have them.These misplaced cells are neurons located unusually in layer one of the cerebral cortex and are devoid of nerve cells 23,25 .Ectopias consist of 50-100 neurons (and glia) that during neural migration, have missed their target in the cortex and escaped into the molecular layer, through a breach in the external glial limiting membrane, accompanied by mild disorganization of the underlying cortical layers which cause a loss of characteristic architectural organization of the cortical neuron (microgyrias and dysplasia).The ectopias and the disordered cortical layering are together referred to as microdysgenesis 25,34 .More so, the number and location of the focal cortical abnormalities vary from brain to brain, affect the language-relevant perisylvian cortex and tend to be more frequent on the left side of the brain than the right 25,34 .Ectopias arise developmentally before the completion of the period of neuronal migration to the neocortex, which in humans takes place between 16 and 20 weeks of gestation 34 .
Ectopias are occasionally seen in routine autopsy studies but only in less than 15 percent of the time, and tend to be few and usually not perisylvian in location 25 .The type of neuronal migration disorder seen in the dyslexic brains may be present in other conditions such as fetal alcohol syndrome and nonspecific mental retardation but those in the latter conditions resulted from obvious injury from the brain during development 25,34 .In the dyslexic brains, on the other hand, it was more subtle and restricted in location 25 .In Galaburda's reports, ectopias were seen in all dyslexic male cases and in one out of two female cases 25,34 .Overall, the dyslexic female brains showed fewer and differently located cortical malformations 24 .These cortical anomalies contain neurons born at different times during histogenesis of the cortex.They lead to alterations in the pattern of connectivity within and between the hemispheres especially with the thalamus, ipsilateral cortex and contralateral cortex and affect the development of these brain areas 33,34 .
An important perspective to the autopsy findings of Galaburda and colleagues [24][25] is that both reduced PT asymmetry and aberrant cortical malformations represent significant neuroanatomical association to developmental dyslexia, even if the symmetry of the brain regions is not classified as pathological and both deviations resulted from abnormal neuronal migration 25 .Neuronal migration refers to a period in brain development when young neurons migrate in search of their final locations in the brain.Neurons are born from neural stem cells in specific proliferative zones far away from the brain areas to their target destination.Neuronal migration, a process which takes place in the cerebral cortex of the brain occurs in an orderly pattern 9,[35][36] and is guided by specialized cells (radial glial cells) that occupy the developing cortex which ultimately gives rise to new neurons through cell division.The termination of migration therefore depends on the integrity of radial cells 9,36 .In some cases, neurons must travel through great distances to reach the cortex.Very tight controls (which are influenced by gene functions) must be in place for neurons to end up in the right place 9,[35][36] .Loss of radial glial cell integrity can therefore cause aberrations in migration anywhere along the neuronal migration path 9,36 .Normal neuronal migration is also dependent on intracellular mechanisms within newly generated neuronal cells that allow for cellular motility.Disruptions in neuronal motility, in radial glia integrity, or in the adhesion between radial glial and new neurons can alter the normal developmental patterning and migration in the neocortex which contribute to establishing abnormal neuronal circuits in brain areas typically devoted to language function 9, 36 .

(iii) Findings in the thalamus
Beside autopsy examinations on the PT, Galaburda and colleagues 27-29 also performed post-mortem examinations on human dyslexics' thalamus.Specifically, the lateral geniculate nucleus (LGN) of the visual pathway 29 , the medial geniculate nucleus (MGN) of the auditory pathway 28 and another study 27 investigated histological changes in the LGN of the primary visual cortex (Broadman area 17).In the LGN, Livingstone et al 29 found that the magnocellular layers of the visual pathway were more disorganized in dyslexic brains and the cell bodies appeared smaller.The decreased magnocellular geniculate neurons have functional consequences while the smaller cell bodies are likely to have thinner axons with slower conduction velocities 29 .The magnocellular system is particu-larly important for the control of eye movements and visual attention 29 .In the auditory system, Galaburda et al 28 reported significantly smaller MGN neurons on the left side compared with the right in the same dyslexic autopsy specimens.No hemispheric asymmetry in MGN neuronal size was observed in ordinary brains.In the primary visual cortex, dyslexic brains exhibited histologic changes in the magnocellular cells of LGN as well as abnormal visually evoked potentials and brain activation to magnospecific stimuli 27 .Overall, the structural deviations found in the LGN of dyslexic brains may be responsible for slowness in early segments of the magnocellular channels, whereas the MGN differences may relate to the auditory temporal processing abnormalities in dyslexia [27][28][29] .

Demerits of autopsy examinations
Although autopsy studies are important in providing direct neuropathological evidence of morphological abnormalities associated with DD and in providing a basis for neuroimaging studies, there are several limitations with the application of such findings.Such limitations include: (i) generalizability is limited due to small sample size (ii) with increased period of storage of autopsy specimens, there is a higher risk of cell shrinkage 37 (iii) reliable identification of microanatomical deviances in general and the boundaries of the PT in particular are difficult to obtain [37][38][39] (iv) all the subjects' historical data was retrospective therefore; it is unclear to what extent brains studied are representative of dyslexic brains in general, (v) autopsy studies are expensive 15,[38][39] and specimens are scarce.
In addition to autopsy examinations, Galaburda and colleagues 33,40,41 also performed series of experiments such as inducing cortical anomalies using animal models.

Animal models
Animal research addresses the question of how minor cortical malformations could lead to clinically persistent disorders of cognitive function 33,40 .As the small size of ectopias precluded the use of imaging techniques, Galaburda and colleagues studied ectopias that had relatively similar morphology to those present in the human dyslexics using animal models 33. 41 .The induction of cortical malformations similar to those found in the dyslexic brain in the rat produced a vari- The South African Optometrist ISSN 0378-9411 196 ety of cognitive deficits 33,41 .

2) Structural imaging studies
The advent of structural imaging techniques (SIT) have helped to address some limitations with the autopsy studies.With the SIT (such as the MRI and CT), measurements of brain structures are performed on living persons.
Four studies [43][44][45][46] found some symmetry of the PT in the dyslexic subjects.The studies include the report by Larsen et al 43 in 1990, which examined the size and symmetry of the PT in group of 19 dyslexic (mean age 15.1 ± 0.3 years) and 19 matched non-dyslexic schoolchildren (mean age, 15.4 ± 0.4 years).The gender ratio was the same in both groups (15 males and four females) and the study was conducted in Norway.Participants were matched for age, intelligence, social-cultural factors, and educational environment.Symmetrical PTs were more common in the dyslexic group (70%) compared with 30% of the control group.The symmetrical PT were attributed to a larger right PT, based on comparing mean lengths of the PT for children with and without symmetrical PT.Furthermore, the authors 43 noted that all subjects with pure phonological deficits in reading had symmetrical PT, which was an indication of neuroanatomical basis for phonological processing impairment in dyslexia.In 1993, a study by Kusch et al 44 in the USA measured the superior surface of the temporal lobe (SSTL) on MRI scans in a sample of 17 dyslexics (Nine males and eight females, mean age, 26.2 ± 15.0 years) and 21 control subjects (Eight males, 13 females, mean age, 33.4 ± 15 years).The SSTL area was divided into anterior and posterior parts (namely, most of the PT and part of the Heschl's gyrus).The authors found that both the anterior and posterior halves of the SSTL area showed significant leftward asymmetry in non-dyslexics, but showed symmetry in dyslexics.According to the authors, this suggests that among dyslexics, the direction of SSTL asymmetry may serve as a risk factor for the severity of reading comprehension problems.Also from the USA, Duara et al 45 in 1991, measured the areas of six bilateral brain segments in the right and left hemispheres, on a horizontal brain section of 21 dyslexics, (twelve males and nine females, mean age, 39 ± 11 years) and 29 non-dyslexic subjects (fifteen males and 14 females mean age, 35.3 ± 10 years) all right-handed.The authors stated that "The region of the brain that includes most of the planum temporale was found to be symmetrical in our study" and then to the contrary stated that "Dyslexic subjects exhibited asymmetry, with the right side greater than the left side".Consequently, this seeming discrepancy has created some inconsistencies in the reporting of the findings by Duara et al 45 , with some authors reporting symmetry [14][15] and others reporting no difference in asymmetry 20,55 .The study by Hynd et al 46 conducted in the USA in 1990 examined the specificity of deviations in patterns of normal brain asymmetry on MRI scans of 10 dyslexics (eight males and two females, mean age, 9.9 ± 2.04 years), 10 with attention deficit disorder/hyperactivity (ADD/H), and 10 (eight males and two females (mean age 11.8 ± 22.0 years) ageand sex-matched control children.Some of the dyslexic participants had attention deficit disorder.The authors 46 concluded that the significant increase in the incidence of PT symmetry seems unique to dyslexia and may be related to deviations in normal patterns of corticogenesis.These findings by Hynd et al 46 correlates with the postmortem findings that 65% of normal adult brains are larger on the left side in the region of the PT reported by Geschwind and Levisky 4 .However, their finding of a significantly smaller than normal left PT in the dyslexic differs from the bilaterally large PT reported at autopsy in dyslexics subjects by Galaburda et al 23 .This finding is important because it indicates that abnormal patterns of symmetry/asymmetry do not occur in all clinical groups of children Although the above reviewed studies [43][44][45][46] all found some symmetry of the PT, the methodological approach across these studies varied.There were differences in slice thickness, imaging plane and the definition of anatomical areas measured as well as differences in subjects characteristics.More importantly, there are also variations in the pattern of symmetry across the studies that reported some symmetry among the dyslexic populations, which therefore precludes a uniform comparison across studies.
The studies that did not find symmetry in the PT among the dyslexic subjects include the study conducted in the USA in 1997 by Rumsey et al 47 , which examined the size and asymmetry of the PT and its extension into the parietal lobe (planum parietale) in the brains of 16 right-handed dyslexic men (all males, mean age, 27 ± 8.0 years) and 14 matched control subjects all males, mean age 24 ± 5.0 years) using the MRI.Approximately 70% to 80% of both groups showed equivalent leftward (left>right) asymmetries of the PT.The authors concluded that given the heterogeneity of the dyslexic population, some subgroup of dyslexic individuals might show unusual symmetry in the PT.From Norway in 1999, Heiervang et al 48 in a population-based study measured the PT and the adjacent planum parietale (PP) region in sagittal MR images of 20 right-handed dyslexic subjects (mean age, 11.80 years) and 20 matched control subjects (mean age 11.75 years).All the participants in this study were males.The results showed a mean leftward PT asymmetry in both the dyslexic and the control group, with no significant difference for the degree of PT asymmetry.Also from Norway and from the same group of researchers Hugdahl et al 49 investigated differences between dyslexic and normal reading children in asymmetry of the PT area in the upper posterior part of the temporal lobe.The participants were 23 dyslexic (nineteen males and four females, the age range indicated was 10-12 years) and 23 normal reading children (twenty males and three females, age range indicated was 10-12 years).Hand preference was indicated.The authors found a significantly larger left than right PT area for both groups.The authors noted that while the right PT area was similar for the dyslexic and control groups, the left PT was significantly smaller in the dyslexic group.In the 1994 detailed study conducted by Schultz et al 50 in the USA that controlled for gender, age and handedness, MRI techniques were used to compare the convolution surface area of the PT, temporal lobe volume and superior surface area of 17 dyslexic children (ten males and seven females, mean age 8.68 ± 0.64 years) and 14 non-dyslexic children (seven males and seven females, mean age 8.94 ± 0.67 years).All subjects were right-handed.According to the authors, the initial analyses suggested smaller left hemisphere structures in dyslexics compared to control subjects whereas subsequent analyses controlling for age, gender and overall brain size revealed no significant differences between dyslexics and non-impaired children on a variety of measures, in particular surface area and symmetry of the PT.Leonardo et al 51 studied anatomical structure (including the PT) in 15 reading disabled (RD) subjects (eight males, seven females, mean age 24 ± 3.0 years) and 15 control (eight males and 7 females, mean age 22 ± 3.0 years) in a volumetric MRI scan.This study was conducted in the USA.The RD group had a more marked leftward asymmetry of the PT, although the group difference was non-significant 51 .Another study conducted by Leonardo et al in 1993 (cited in Leonardo et al) 51 also found a leftward asymmetry of the PT.Robicho et al 52 in 1999 measured cortical asymmetries of posterior language-related areas, including the PT in 16 adult male dyslexic subjects (mean age 21 ± 0.2 years) and 14 age-matched male controls (mean age 23.6 ± 3.9 years) in France.Nine of the 16 dyslexic subjects were right-handed and four of the controls were not.They found no differences in PT asymmetry between the two groups and concluded that phonological segmentation skills may relate to frontal lobe morphology, while phonological memory-based impairment in people with dyslexia may rather relate to parietal lobe asymmetry.Using similar methods as in the study by Robichon et al 52 another study by the same group of researchers, Habib and Robichon 53 also found no differences in PT asymmetry.The study by Best and Demb 54 presented another perspective.The authors measured the PT in five dyslexic subjects (three males and two females, mean age 22 ± 2.9 years) with a documented magnocellular deficit and five controls subjects (Three males, two females mean age 26.8 ± 6.1 years).The study was reported in 1999 and all the dyslexic subjects showed normal, leftward asymmetry of the PT 54 .Best and Demb 54 concluded that planar symmetry may be associated with a subgroup of dyslexia.
The reviewed studies 47-54 that did not find symmetrical PT in dyslexic population also has methodological variations across the studies.Despite a similar report of non-symmetrical PT, the findings of these studies differed in some ways, for example, Heiering et al 48 and Hughal et al 49 found leftward PT asymmetry in both the dyslexic and the control group with the dyslexics having smaller left PT.Schutz et al 50 found no significant differences between dyslexics and nondyslexic on a subsequent analysis while Leonardo et al 51 found exaggerated leftward asymmetry of the PT in the dyslexic group.
Taken together, a trend with autopsy and MRI studies on the dyslexic population is that results appeared to be consistent across groups of researchers.For example, Galaburda and colleagues [21][22] consistently found symmetrical PT in all their postmortem subjects.Leonarndo et al 51 study found exaggerated leftward asymmetry.Robicho et al 52 and Habib and Robichon 53 reported no differences in PT asymmetry between dyslexics and non-dyslexics.This tendency may be an indication of some methodological bias.

Discussion
Autopsy studies on dyslexia have been conducted by the same group of researchers (Galaburda et al)  and the results are fairly consistent with all subjects examined having symmetrical PT and several with cortical malformation, mainly ectopias and dysplasia 21,[23][24] .However, results from the MRI studies reviewed on PT [43][44][45][46][47][48][49][50][51][52][53][54] have shown an inconsistent trend, for example, the autopsy findings of Galaburda and colleagues have been replicated only in a few of the MRI studies [43][44][45][46] .The inconsistencies in the findings reported by different researchers have been attributed to methodological differences across studies 14,50 .The limitations may be broadly classified as: issues related to measurements and technique (definition of anatomical borders, dimensions measured, slice thickness, measurement plane) as well as subjects' characteristics (gender, age, handedness, sample size and differences in dyslexia diagnostic standards, and co-diagnoses) as well as intelligence quotient.These limitations are discussed below.
Specifically, a major difficulty in studies conduct-ed on PT is related to the definition of its anterior and posterior borders 14,55,56 .Studies 14,56,57 have shown that the way the anatomical borders of the PT is defined affects the measurement outcome.Zetzche et al 56 have outlined different methods of description of the PT and remarkably, studies on PT have used different definitions of the PT.In particular, there has been confusion as to exactly what the PT is and whether the second Heschl's gyrus, (if present) should be included and which one of the planes created by the posterior branching of the Sylvian fissures constitutes the continuation of the PT 58 .Therefore, issues related to the inconsistencies in the definition of the PT has to be resolved before results by the different researchers can be appropriately compared 14,56,58 .The difficulty of determining the exact anatomical definition of the PT also led to some authors measuring some indirect structures, as in the study by Kusch et al 44 .Also, inter-measurer reliability in the measurements of the PT is poor 55 .Another possible source of inconsistent results in PT research relates to the dimension of PT measurements.Invariably, the size of structures of the brain can be estimated using linear, area, or volumetric measurements 14 and different studies have used different dimensions.The use of any measurement dimension, which does not account for the convoluted nature of the PT, is likely to obtain unreliable results 14,60 .
Furthermore, different MR imaging plane (axial/ horizontal, coronal or sagittal) have been used in different studies to measure shape, size and asymmetry of the PT.According to Galaburda 58 although each imaging plane has unique advantages, there are also limitations with individual technique.For example, a demerit with the commonly employed horizontal (axial) plane in MRI is that it is too close to the plane of the PT itself, making visualization difficult even for experienced investigators 14,58 .Related to the problem of imaging plane is that different MRI slice thicknesses (Hynd et al 46 =7.5 mm, Larsen et al 43  One of the factors related to subjects' characteristics is gender.Some MRI studies combined the results from disproportionate ratios of male and female subjects and controls despite evidence that gender is an important variable in brain morphology.Support for SO Wajuihian -Neurobiology of developmental dyslexia: Part 1: ... autopsy and structural neuro-imaging studies The South African Optometrist ISSN 0378-9411 199 the hypothesis that brain language areas may be differently organized in males and females comes from autopsy studies 24 , MRI studies 17, 61-62 and a functional magnetic imaging (fMRI) study by Shaywitz et al 63 who demonstrated gender differences in the localization of brain activation during phonological processing.In addition, girls and boys differ in language development ability 64 and that reading disability affect boys more frequently than girls, with reported gender ratios 65-66 of 3:1.Also, on average, men have 8-10% larger brains than women, therefore the size of a particular brain structure, such as the PT, can also be expected to be larger in men 16 .Examples of studies with disproportionate gender ratio are the study by Heivering et al 48 , where all participants from both groups were males and the study by Hugdahl et al 49 where subjects from the dyslexic group consisted of 19 boys and four girls and 20 boys and three girls for the control group.
Another methodological variable related to subject characteristics is handedness (dominant hand/hand preference).Determining the subjects' hand preferences is important because handedness is believed to relate to functional language lateralization and individuals with consistent right-handedness have been reported to show more pronounced asymmetries of the PT 14,16,47,56,[67][68] .Strict controls for handedness are essential in anatomical studies as failure to do so is a major methodological flaw 32 .
Finally, other methodological factors, which may contribute to differences in findings across studies, include age, small sample sizes, criteria used to define dyslexia, the heterogeneity of the disorder, and co-diagnoses (such as attention-deficit/hyperactivity disorder, specific language impairment).Age as a variable may constitute a confounder due to the effects of brain development which relates to a possible difference between child and adult studies 14 .The samples sizes across studies are typically small with an average of 15 for dyslexic subjects for the studies reviewed.Small sample sizes do not have a high statistical power to detect differences in variables such as handedness, age and gender differences.For example, the study by Schultz et al 50 demonstrated a significant influence of gender and age on brain morphology.In addition, diagnostic criteria for dyslexia varied across studies [43][44][45][46][47][48][49][50][51][52][53][54] .

Recommendations for future studies
Given the outlined limitations with the studies reviewed, different authorities have recommended various methodological modifications aimed to improve the consistencies from neuro-imaging studies.These recommendations are summarized below.

1)
A clear operational definition of the anatomical borders of the PT must be established and used across studies.

2)
Slice thickness ranging between 1.5-2.0mm which will allow good visualization of cortical foldings and enables accurate identification of the PT boundaries is recommended 14,32 .3) Coronal rather than sagittal imaging sequences are preferred because these yield more distinct image slices through the PT.Also since the PT is a convoluted three-dimensional structure, no single plane is optimal for display of the PT.Only the volumetric MRI acquisitions allows the use of a computer programme that allows simultaneous display in all three orthogonal views, in the axial, sagittal and coronal planes which substantially improves anatomical border determination and has been suggested by some studies 14, 16, 58 . 4) Sample sizes must be large enough to allow a high statistical power to adequately address variables such as handedness, age and gender differences 69 .

5)
Handedness, age and gender difference should be adequately controlled.Accurate measurement, preferably objective and subjective of handedness is suggested 14 .
The possible limitation with this review is that the literature search is limited to English-only sources.However, the major strength is that it is a thorough and up-to-date review paper on the neuro-anatomy of developmental dyslexia.

Summary and Conclusion
Overall, only the autopsy investigations conducted by Galaburd et al 21,[23][24] have been consistent in their findings of unusual symmetry of the PT, cortical malformations and differences in the thalamus.The reports from MRI studies were inconsistent possibly due to methodological variations across the studies.

The South African Optometrist
ISSN 0378-9411 200 Although the occurrence of symmetrical PT may not confirm the development of dyslexia, the occurrence of symmetrical PT in several consecutively autopsied subjects may not simply be due to chance.Also, given that these same findings (PT symmetry) have been corroborated by some MRI studies suggests that symmetrical PT may be relevant in the etiology of DD.However, the fact that there are also variations in the pattern of symmetry reported across the studies that found some symmetry of the PT (in MRI studies) among the dyslexic populations makes it difficult to make a firm conclusion in support of symmetrical PT in dyslexia.Therefore, given the available evidence from the reviewed studies, no anatomical structure appears to firmly characterize dyslexic individuals.The neurobiology of dyslexia remains a subject of continued research and it will be important to investigate the relevance of neuro-anatomical structures in the etiology of dyslexia.Consequently, Galaburda 25 reported that the presence of enlarged PT associated with symmetrical PT may be related to increased number of axons passing through the corpus callosum (nerve fibre connecting both hemispheres of thee brain) which then creates an anomalous interhemispheric pathways to the perisylvian-language regions.Due to its role in interhemispheric transfer, the corpus callosum has been another area of intense research in the quest for the neurobiology of DD.Evidence from MRI studies of the CC will be reviewed in part 2 of the series on the neurobiology of development dyslexia.This review is intended to contribute to the understanding of dyslexia and to guide future research.
relatively specific to dyslexia 46 .