Imaging biomarkers of outcome in the developing preterm brain

Review

Imaging biomarkers of outcome in the developing preterm brain Laura R Ment, Deborah Hirtz, Petra S Hüppi Lancet Neurol 2009; 8: 1042–55 Published Online October 1, 2009 DOI:10.1016/S14744422(09)70257-1 Departments of Pediatrics and Neurology, Yale University School of Medicine, New Haven, CT, USA (L R Ment MD); National Institute for Neurological Disorders and Stroke, Bethesda, MD, USA (D Hirtz MD); and Department of Pediatrics, Children’s Hospital, University Hospitals of Geneva, Geneva, Switzerland (P S Hüppi MD) Correspondence to: Laura R Ment, Department of Pediatrics, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06520, USA [email protected]

1042

The neurodevelopmental disabilities of those who were born prematurely have been well described, yet the underlying alterations in brain development that lead to these changes remain poorly understood. Processes that are vulnerable to injury in the developing brain include maturation of oligodendrocyte precursors and genetically programmed changes in cortical connectivity; recent data have indicated that diffuse injury of the white matter accompanied by neuronal and axonal disruption is common in prematurely born infants. Recent advances in MRI include diffusion tensor imaging and sophisticated image analysis tools, such as functional connectivity, voxel-based morphometry, and mathematical morphology-based cortical folding strategies. These advanced techniques have shown that white matter structure is dependent on gestational age and have started to provide important information about the dynamic interactions between development, injury, and functional recovery in the preterm brain. Identification of early biomarkers for outcome could enable physicians and scientists to develop targeted pharmacological and behavioural therapies to restore functional connectivity.

Introduction Preterm birth is among the leading public health problems in the USA and Europe.1 There are more than 4 million livebirths in the USA every year, and more than 1∙5% of these infants weigh less than 1500 g (ie, have a very low birthweight). Survival for these medically fragile neonates is about 70–80%, but the annual new perinatal care costs for these infants exceeds US$18 billion every year in the USA alone.1–5 Cerebral palsy has long been recognised as a frequent outcome of premature birth, but substantial cognitive, behavioural, and other neuromotor abnormalities are also being recognised at school age and beyond.6–11 Preterm infants are at high risk of brain injury, yet the dynamic interactions between the genetically determined programme of development, injury, and functional recovery remain poorly understood. MRI has enabled non-invasive high-resolution evaluation of the developing brain, and many studies have documented the macrostructural sequelae of premature birth, including delayed cerebral grey–white matter differentiation and diffuse excessive hyperintense signal in the developing white matter.12–17 Furthermore, when compared with term control infants, preterm infants have global and regional decreases in cortical grey and deep grey matter, less myelinated white matter, smaller corpus callosal areas, and substantial ventriculomegaly, yet the long-term neurodevelopmental predictive value of these relatively large global and regional volumetric changes remains limited.18–24 Cerebral white matter injury, or the periventricular leukomalacia (PVL) complex, has long been recognised as the most common neuropathology in prematurely born infants, and one crucial MRI finding is that of decreased white matter volumes at term-equivalent age. White matter injury in preterm neonates is accompanied by diffuse neuronal and axonal disease affecting not only cerebral white matter but also deep grey matter, cortical, and cerebellar areas.25 During this crucial period of human brain development, cortical folding occurs, and

little is known about the effect of cerebral white matter injury on this important process.26,27 Imaging strategies that provide data on both connectivity and cortical development might reveal important early biomarkers that are predictive of later development. Two relatively new techniques, diffusion tensor imaging (DTI) and functional MRI connectivity (fcMRI), combined with advanced image analysis tools, such as voxel-based morphometry (VBM) and mathematical morphologybased analysis of cortical folding, provide complementary data for understanding development, injury, and recovery in the developing brain. These non-invasive imaging techniques might guide our understanding of alterations in brain development and their correlation with functional outcome in prematurely born infants.

Preterm birth and disability Data from studies from the USA, Europe, and Australia have all revealed poorer educational outcomes in preterm children than in term controls. At school entry, minor developmental impairment is diagnosed in 30–40% of preterm children and major disabilities in almost 20% of preterm children.2,8,9,28,29 More than half of these children require special assistance in the classroom, 20% are in special education, and 15% have repeated at least one grade in school.30,31 In France, 42% of children born at 24–28 weeks of gestation and 31% of those born at 29–31 weeks needed special health-care support owing to neurological sequelae, compared with only 16% of those born at 39–40 weeks.8 The intellectual deficits of preterm subjects can persist through adolescence and young adulthood.32 In a study by the Victorian Infant Collaborative Study Group, 14% of extremely low birthweight preterm children (500–999 g) were classified as severely disabled and 15% were classified as moderately disabled at the age of 14 years, compared with 2% of controls who were classified as severely disabled at this age.33 Similarly, when 1097 Dutch preterm children with a gestational age of less than 33 weeks and/or a birthweight of 1500 g were assessed at the age of 14 years, more www.thelancet.com/neurology Vol 8 November 2009

Review

than half had disorders in learning, attention, and social–emotional skills.34 Finally, when compared with healthy controls, fewer prematurely born young adults continue their education after high school, and fewer have full-time employment or live outside the parental home as adults.35,36

Injury in the developing preterm brain Injury is common in the preterm brain.25 Although parenchymal involvement of intraventricular haemorrhage and focal PVL have been viewed as the major forms of brain injury that cause cerebral palsy after preterm birth, intraventricular haemorrhage with periventricular haemorrhagic infarction now occurs in about 5% of infants with very low birthweight, whereas cases of focal PVL have decreased in recent years. This is not the case for the more diffuse injury in the periventricular white matter of neonates with very low birthweight. Results from conventional MRI studies indicate that up to 50% of neonates with very low birthweight have signs of such diffuse white matter injury and accompanying neuronal or axonal disease.25 First described by Banker and Larroche,37 PVL is a unique pattern of neonatal brain injury involving not only the cerebral white matter but other areas as well. PVL is characterised by both focal macroscopic and microscopic regions of necrosis, loss of premyelinating oligodendrocytes, and marked astrogliosis and microgliosis of the white matter; for many neonates, cerebral ventriculomegaly detected with cerebral ultrasonography is the only biomarker of this cascade of events. The axonal and neuronal changes found in neonates with diffuse white matter injury are less well described.38–40 MRI studies of prematurely born subjects from the neonatal period to late adolescence show decreased volumes of cortical grey matter and alterations in the thalamus, basal ganglia, and cerebellum.16,21,41–44 As the late second and third trimesters are the time of both axonal ingrowth and elaboration of synaptic connections, and oligodendroglia produce guidance molecules in addition to myelin,45 microstructural changes in connectivity have been postulated. In addition to MRI findings and neuropathological results, preclinical studies indicate the occurrence of both oligodendroglial and neuro-axonal injury.46–49 Both hypoxic– ischaemia and chronic hypoxia cause acute degeneration of late oligodendroglia precursors, followed by a regenerative response of the surviving preoligodendroglia. However, the surviving cells have arrested maturation with failure to generate myelin.50 Several authors have noted regional vulnerability of the developing brain on imaging studies, and, in these preclinical studies, the regional extent of white matter damage coincides with the presence of susceptible populations of late oligodendroglia.48,51

In vivo imaging strategies DTI and blood oxygen level dependent (BOLD)-based functional MRI, combined with advanced image analysis tools such as automatic voxel-based analysis methods for www.thelancet.com/neurology Vol 8 November 2009

whole-brain comparisons, provide complementary in vivo data for understanding the effect of preterm birth on the connectivity and functional competence of the developing brain. Additionally, recent assessments suggest that primary cortical folding might provide an early marker of later behavioural development and might indicate microstructural changes that affect functional cortical field development and neural connectivity. These strategies are briefly reviewed in the panel and are more extensively described in the webappendix. DTI measures water diffusion in biological tissues at a microstructural level and is a powerful technique to study the structural basis of white matter development.52 As such, this technique provides a unique imaging tool to assess important microstructural changes in brain development, such as axon calibre changes and premyelination and myelination stages, all likely to be affected by prematurity. DTI parameters include the three diffusion tensor eigenvalues (λ1, λ2, λ3, which represent diffusion along the three tensor principal axes), the mean diffusivity (Dav) or apparent diffusion coefficient (ADC), and a mathematical measure of anisotropy, which describes the degree to which water diffusion is restricted in one direction relative to all others (panel and webappendix). Fibre tracking uses each voxel’s primary eigenvector of the diffusion tensor to follow an axonal tract in three dimensions from voxel to voxel through the brain, thus enabling the delineation of specific cerebral white matter tracts and connectivity. Different tractography tools have been developed; the most common are streamline deterministic and probabilistic fibre tracking.53 There are several approaches to analysing DTI data, including region of interest approaches, VBM approaches,54–56 histogram analyses, and tract-based analysis57 (see webappendix and elsewhere for reviews57–59). It is likely that the functional deficits associated with prematurity are partly due to alterations in cortical connectivity. The different image analysis tools that enable the three-dimensional visualisation of brain fibre tracts (tractography) provide the opportunity to study and understand the structural alterations underlying functional deficits. Diffusion imaging enables the study of the establishment of brain connectivity and plasticity during a time period of extreme importance for the structural and functional integrity of the brain. New diffusion-based approaches (such as AxCaliber60) for the measurement of axonal diameter and diffusion spectrum imaging61,62 to characterise establishment of corticocortical connections during development will further refine image-based knowledge on brain development and integrity (figure 1, webvideo). Although there are only a few studies using fcMRI, a method to assess both neural processing and resting state connectivity, in newborns,63 this technique might also offer insights into the microstructural changes in brain development that occur in preterm infants.64,65 In resting state fcMRI, bandpassed-filtered (0∙012–0∙1 Hz) BOLD signal intensity time courses show coherent spontaneous

See Online for webappendix

See Online for webvideo

1043

Review

oscillations between specific areas of the brain, and this new technique has been applied to study interactions between spatially distinct regions in typically developing children and adolescents.66,67 Research into resting states suggests that functional connectivity is at least partially anatomically determined in typically developing subjects. Panel: Glossary of imaging terms Diffusion tensor imaging (DTI) is an MRI method for which the image intensities at each position are attenuated depending on the strength (b-value) and direction of the so-called magnetic diffusion gradient, as well as on the local microstructure in which the water molecules diffuse. This method enables the measurement of the restricted diffusion of water in tissue and allows the study of microstructural changes and structural connectivity. Apparent diffusion coefficient (ADC) is a numerical measure of diffusion. The ADC in anisotropic tissue varies depending on the direction in which it is measured. Diffusion is fast along the length of (parallel to) an axon and slower across it perpendicularly. For such anisotropic tissue, diffusivity is best described by the diffusion tensor ellipsoid, defined by three eigenvectors and their lengths (eigenvalues, with λ as a measure of diffusion along each of the three primary axes). Mean diffusivity (Dav) is the sum of the diffusivity along the principal axis (λ1; also called the longitudinal diffusivity or the axial diffusivity) and the diffusivities in the two minor axes (λ2 + λ3; a measure of radial diffusivity) divided by three. Relative anisotropy and fractional anisotropy are measures of relative degree of directionality of diffusion in a voxel. The basis of these calculated parameters is the sum of squares of the diffusivity differences. The values vary between 0 (isotropic) to 1 (anisotropic). Vector maps, or RGB colour-coded directionality maps, indicate the orientation of the major eigenvector of the diffusion tensor, and are typically overlaid on structural images. Diffusion tractography is a general term indicating various methods of reconstructing white matter pathways in vivo, based on the assumption that the principal direction of diffusion in white matter is parallel to the main fibre direction in every voxel. Voxels are units of tissue volume that correspond to the smallest three-dimensional space from which MRI data are acquired. Computational image analysis strategies such as voxel-based morphometry, deformation-based morphometry, and tract-based spatial statistics are used to assess group-wise regional microstructural morphological differences in both grey and white matter, based on input from conventional MRI or DTI (see supplementary material for further detail). Functional MRI (fMRI) technically refers to the use of the non-invasive MRI technology used to detect regional changes in signals that are correlated with brain functional activity. fMRI uses deoxygenated haemoglobin as an endogenous contrast agent to indirectly depict cortical activated regions (therefore, called blood oxygenation level dependent contrast; BOLD). Functional MRI connectivity (fcMRI) refers to low-frequency fluctuations of the BOLD fMRI signal between spatially distinct regions; it can be either task-related or can indicate resting state connectivity. Mathematical morphology, a novel technique for the analysis and processing of geometrical structures and shapes, is used to assess primary cortical folding with measurement of surface area and cortical gyration and sulcation index, through curvature measurements. These measures are dependent on both cortical grey and underlying white matter connectivity. Partial-volume averaging refers to the loss of contrast between two adjacent tissue types caused by insufficient spatial resolution (eg, the cortex–CSF interface in deep sulci).

1044

This finding is of particular importance to preterm infants in whom white matter development might be compromised.68–70 There can be several design problems with fMRI, including registration of infants’ brains, selection of BOLD signal significance values, and analysis strategies (see webappendix and elsewhere for review 71). Finally, computational approaches using mathematical morphology, a novel technique for shape processing and analysis, can be used to investigate primary cortical folding in the preterm brain. Through three-dimensional reconstruction of the interface between the cortex and white matter, this approach quantifies both surface area and gyration through curvature measurements.72,73 To accurately define the cortical surface, one would ideally identify the cortical grey matter–CSF interface, but partialvolume averaging precludes this with current image resolution. Instead, the interface between the cortical grey matter and the white matter is reliably identified by use of a mathematical morphology approach (see webappendix and elsewhere for review;74 figure 2). As the incidence of handicap increases among those who are prematurely born, the need for early biomarkers of injury and outcome is crucial to the care of this vulnerable population. These strategies have the potential to provide important early information about the effect of preterm birth on structure and function in the developing brain. Nonetheless, there are many challenges faced by those who use and interpret these strategies that might affect reliability of the data. Understanding the difficulties that are unique to each technology is crucial to our understanding of the data that they produce (webappendix).

Imaging the preterm brain: the neonatal period During the late second and third trimesters of gestation, cerebral development is characterised by sequential periods of cellular proliferation, migration of glia and neurons into appropriate cortical positions, axonal ingrowth, and the elaboration of synaptic connections. Although most cortical neurons have migrated before this time, neurogenesis continues in the superior ventricular zone during the second trimester and the population of subplate neurons attains maximal development at 24–32 weeks. Finally, maturation of the oligodendroglial lineage and the initiation of myelination are the hallmark features of the third trimester.

Microstructural development of the white matter The diffusion and anisotropy parameters of developing white matter have been well described.75–85 ADC values of white matter are directly associated with gestational age75 and, during the third trimester of gestation, ADC values of cerebral white matter decrease, whereas those for anisotropy increase.77,79 Decreases in diffusion are observed principally in λ2 and λ3, which reflect alterations in water diffusion perpendicular to white matter fibres and indicate changes due to both pre-myelination and myelination.78,86 Changes during pre-myelination have been attributed to www.thelancet.com/neurology Vol 8 November 2009

Review

an increase in the number of axonal microtubule-associated proteins or calibre, as well as to the presence of oligodendroglia. When healthy preterm infants were compared at termequivalent age with term controls, there was no difference in ADC values between the groups (table 1).76 By contrast, fractional anisotropy (FA) values for central white matter regions, including the posterior limb of the internal capsule, the centrum semiovale, and genu of the corpus callosum, were all significantly lower in the preterm group. Those infants with the lowest gestational ages (<28 weeks) had additional decreases in FA.80,85 Data from both fibre-tracking investigations of preterm neonates of varying gestational ages and serial tractography studies indicate that FA values in corticospinal tracts are significantly correlated with postmenstrual age (ie, gestational age plus time elapsed after birth).79,81 As these tractography methods use MRI indices from whole-brain white matter, without inter-subject registration, the anatomical localisation is expected to be more accurate than in studies that use automatic voxel-based comparisons. By contrast, recent data suggest that in certain regions of the brain, FA values for preterm infants at term-equivalent age might exceed those for term control infants. In infants who had neonatal ultrasounds free of major structural lesions, Gimenez and colleagues82 noted higher values for FA in the inferior frontal occipital fascicle, stria terminalis, fornix, optic radiations, inferior longitudinal fascicle, and lateral geniculate nucleus in preterm infants at termequivalent age when compared with term controls. These areas correspond to fibre tracts of the neurosensory pathways of vision and hearing, functions that mature more rapidly in preterm infants owing to early experience. Similarly, Rose and colleagues83 found significantly higher FA values in the corticospinal tract at the level of the cerebral peduncles of preterm infants (25–29 weeks’ gestational age) studied at term-equivalent age when compared with term control infants. Rose and co-workers examined term infants at less than 3 days after birth and found that their T2 measurements had higher water content than those of preterm infants, possibly consistent with physiological data suggesting increased neonatal water content at birth. These authors concluded that, in an area of similar fibre tract maturation between preterm and term infants (ie, the cerebral peduncle), higher water content caused altered diffusion parameters in term infants compared with preterm infants at term-equivalent age.

Effects of injury on neonatal white matter microstructure The early assessment of white matter in preterm infants with diffusion imaging can reveal bilateral periventricular diffusion restriction similar to the typical distribution of focal PVL when ultrasound and conventional MRI show no or non-specific abnormalities.22,84 Histological changes in the acute phase of PVL alter white matter microstructure and water diffusivity, and a reduced ADC in the periventricular white matter in an otherwise normal www.thelancet.com/neurology Vol 8 November 2009

Figure 1: Network of structural connectivity established from DTI in a newborn infant This DTI-based map of global white matter fibres is the basis for the extraction of a whole-brain structural connectivity network already established for the adult brain.61 These networks enable definition of the connection matrix, strength of connectivity dependent on fibre density, connector hubs, and structural networks core, all of which are tightly linked to functional networks. Colours indicate the direction of the fibres according to convention: red=left-right; green=anterior-posterior; blue=up-down. Figure generated in collaboration with Petra Hüppi, Elda Fischi, Leila Cammoun, Jean-Philippe Thiran (Ecole Polytechnique Federale de Lausanne, Switzerland), Patric Hagmann, and Reto Meuli (Centre Hospitalier Universitaire Vaudois Lausanne, Switzerland). DTI=diffusion tensor imaging.

A

B

C

Figure 2: An example of brain surface rendering from a preterm infant This figure illustrates the development of cortical folding in the preterm brain derived using a mathematical morphology approach. Images taken at 26 weeks’ gestation (A), 29 weeks’ gestation (B), and 36 weeks’ gestation (C). The colours used outline the surface curvature. Images generated in collaboration with Petra Hüppi, Jessica Dubois, and Jean-Francois Mangin (Neurospin, Saclay, France).

preterm brain is considered an early indicator of white matter injury.58 Anisotropy of white matter also changes after focal and diffuse injury. In the chronic stage of PVL, decreased relative anisotropy can be present and vector maps might show disruption of white matter tracts distant from focal lesions detected on conventional imaging.76 In this case, changes in anisotropy are found not only near the site of primary injury, but also in the posterior limb of the internal capsule, indicating a disturbance of developing fibres that project through this area. In this way, anisotropy and vector maps show injury that is not detectable by more conventional means.58 For preterm infants with moderate white matter injury, the expected progression of ADC and FA values at termequivalent age is not seen. Miller and colleagues77 noted 1045

Review

increasing ADC values with gestational age in both frontal white matter and occipital visual regions in preterm infants with white matter injury, although the expected changes in anisotropy values with maturation did not occur, indicating disruption of normal brain maturation. Finally, in preterm neonates with no evidence of white matter injury on traditional MRI at term-equivalent age, Anjari and colleagues85 recently reported that, independent of the degree of prematurity, respiratory disease was associated with cerebral white matter abnormalities.

White matter injury in subcortical structures Boardman and colleagues44 tested the hypothesis that preterm infants with diffuse white matter injury would have substantial cerebral morphological alterations when compared with preterm infants without white matter injury and term controls. All infants were studied at termequivalent age, and the data indicated reduced thalamic and lentiform volumes in preterm infants compared with term controls. These changes increased with decreasing gestational age for infants who were born at less than 28 weeks of gestational age and were more common in infants with diffuse white matter injury.44 There was no a priori hypothesis about the specific regions affected and the methods were based on a whole-brain approach similar

to VBM techniques. These data provided an estimate of voxel-wise volume change for all transformed images compared with a standardised control template image; this sophisticated rater-independent method was subsequently confirmed in a second study by manual segmentation.87

Cortical folding Several authors have investigated the effect of preterm birth on primary cortical folding88–90 because white matter connectivity has the potential to partly determine gyrification in the developing brain.27 Several hypotheses have been suggested to explain the processes that underlie gyrification in the developing brain, including genetic control, active growth of convolutions during gyrogenesis, differential growth of inner and outer cortical layers, cortical growth, and tension from white matter axonal fibres. All these factors might be affected by preterm birth and associated injury.26,27,91,92 Biagioni and colleagues93 suggested that, for preterm neonates, cortical folding significantly increases with postmenstrual age. Both Dubois and colleagues72 and Vasileiadis and co-workers94 found smaller cortical and white matter volumes but equivalent sulcation in female preterm infants compared with male preterm infants. Preterm neonates with intrauterine growth restriction had

Number

Birthweight/ Major exclusion criteria GA

Age at MRI

Outcome measures

Results

Hüppi et al75

17 preterm, 7 term

25–35 weeks

NA

Serial studies over first 2 weeks after birth and at term

ADC and RA; ROI

RA directly associated with gestational age Preterm infants had lower RA in internal capsule (p<0·01) and central WM (p<0·01) than did term infants

Hüppi et al76

10 preterm with WM disease or injury, 10 preterm controls

29 weeks

NA

No difference in ADC between groups at term-equivalent age Serial studies over ADC, RA of central, Infants with WM injury had significantly lower RA in central first 3 weeks after birth anterior frontal, and at term occipital, and temporal WM (p=0·03) and PLIC (p=0·02) than did control infants WM, and PLIC; ROI

Miller et al77

25–34 weeks 11 preterm normal, 7 preterm with mild WM disease or injury, 5 preterm with moderate WM disease or injury

Serial studies over Congenital infection, 2–6 weeks after birth malformations, parenchymal haemorrhage and at term and infarctions

ADC and anisotropy; ROI

In normal infants, ADC decreased and anisotropy increased with age In infants with moderate WM disease or injury, ADC did not decline and anisotropy did not increase

Partridge 14 preterm et al78

25–34 weeks

WM injury on MRI, ≥grade 2 IVH, congenital infection, malformations

First scan at 28–39 weeks’ postmenstrual age; 8 of 14 were serially imaged at near-term age

Tract-specific maturation of WM: deterministic and probabilistic tractography

Commissural tracts of CC and internal capsule matured earlier than in subcortical projection and association pathways Maturation characterised by increasing FA and decreasing mean diffusivity

Berman et al79

37 preterm

NA

IVH, ventriculomegaly (patients with minimal WM lesions included)

28–43 weeks’ postconceptual age; 10 with serial scans

FA, transverse diffusion of motor and sensory tracts: deterministic and probabilistic tractography

All tract-specific diffusion parameters significantly correlated with age Motor tracts had higher FA and lower diffusivity than sensory tracts

Anjari et al80

26 preterm with normal MRI, 6 term controls

25–33 weeks

Focal lesions on MRI

Term

TBSS and FA

Preterm infants had lower FA in centrum semiovale, frontal WM, and genu of CC (p<0·001 for all) Infants with <28 weeks’ gestational age had additional decreases in FA in external capsule, PLIC, isthmus, and midbody of CC (p<0·05 for all)

Dudink et al81

28 preterm with normal neonatal MRI and normal development at 1–3 years of age

25–32 weeks

Congenital infections and malformations, WM injury, IVH, or ventriculomegaly on MRI

First 4 days after birth

ADC and FA of WM tracts; ROI and seed point tractography

Significant correlation between gestational age and FA of PLIC (r=0·495; p<0·01)

(Continues on next page)

1046

www.thelancet.com/neurology Vol 8 November 2009

Review

Number

Birthweight/ Major exclusion criteria GA

Age at MRI

Outcome measures

Results

(Continued from previous page) Gimenez et al82

27 preterm, 10 term

28–34 weeks

Neonatal ultrasound with Term major lesions (2 preterm with germinal matrix haemorrhage and 1 with ventriculomegaly included)

FA of WM tracts and voxel-based morphometry

Preterm infants had higher FA in inferior fronto-occipital fasciculus, stria terminalis, fornix, optic radiations, inferior longitudinal fasciculus, and lateral geniculate (p<0·001 for all)

Rose et al83

12 very preterm, 11 preterm, 10 term

25–29 weeks 29–32 weeks

Cortical or WM injury or haemorrhage

Term

FA of WM tracts and TBSS

Term infants had higher FA in splenium of CC and corona radiata than did preterm infants Preterm infants had higher FA in cerebral peduncle than did term infants Preterm infants had higher FA in frontal WM, genu of CC, external capsule, centrum semiovale, and sagittal striatum than did very preterm infants Very preterm infants had higher FA in cerebral peduncle than did term infants

Cheong et al84

111 preterm

<1250 g and/or gestational age <30 weeks

NA (13·5% infants had IVH, 4·5% had cystic periventricular leukomalacia)

Term

ADC, FA, and axial and radial diffusivity; ROI

39 infants had normal, 59 had focal, and 13 had extensive WM signal intensity abnormalities on MRI Infants with extensive WM signal intensity abnormalities had less FA in interior capsule, right inferior frontal gyrus, and right superior occipital regions, and higher radial diffusivity in the right internal capsule, bilateral sensorimotor, and right superior occipital regions. These infants also had lower ADC in bilateral sensorimotor and right superior occipital regions, and lower axial diffusivity in bilateral sensorimotor (p<0·05 for all)

Anjari et al85

53 preterm

24+2 to 32+4 weeks

Focal lesions on MRI

Term

TBSS and FA

FA linearly associated with gestational age at birth Independent of gestational age, infants with mechanical ventilation for >2 days (n=10) had significant decreases in FA in the genu of CC, and infants with chronic lung disease (n=15) had significantly lower values of FA in the left inferior longitudinal fasciculus

ADC=apparent diffusion coefficient. CC=corpus callosum. DTI=diffusion tensor imaging. FA=fractional anisotropy. GA=gestational age. IVH=intraventricular haemorrhage. NA=not available. PLIC=posterior limb of internal capsule. RA=relative anisotropy. ROI=region of interest. TBSS=tract-based spatial statistics. WM=white matter.

Table 1: Effects of preterm birth on WM maturation in neonates—DTI studies

a more pronounced decrease in volume in relation to surface area and increased sulcation; these values correlated with impaired behavioural functions.72 Additionally, preterm infants with early white matter lesions were more likely to have increased gyrification in overlaying cortex that might lead to overt polymicrogyria or increased sulcation.73,95 These abnormalities in cortical folding have been associated with functional development in preterm infants at term-equivalent age; they might be a marker for those early changes in cortical development that give rise to the large percentage of preterm infants with cognitive impairment in the absence of focal cerebral lesions.73 Ramenghi and colleagues88 quantitatively assessed brain development in preterm infants with MRI-diagnosed white matter injury and matched preterm controls with a normal MRI appearance at term-equivalent age. Myelination and cortical folding were significantly delayed in infants with white matter injury. Therefore, similar to FA and thalamic volume measurements, cortical folding is altered by white matter injury.

Structure–function relationships Central white matter DTI parameters obtained in preterm neonates at term-equivalent age might provide www.thelancet.com/neurology Vol 8 November 2009

useful prognostic data as cerebral palsy and deficits in gross motor function have long been attributed to alterations in central white matter development in preterm infants (table 2).96,97 Arzoumanian and colleagues96 evaluated DTI studies at term-equivalent age and follow-up neurological examinations at a corrected age of 18–24 months (age corrected for prematurity) on 137 preterm neonates, and found that FA values in the right posterior limb of the internal capsule were significantly lower for preterm infants with cerebral palsy compared with those with normal examinations. For preterm children who underwent developmental testing at 18–24 months’ corrected age, the Bayley psychomotor developmental index correlated with FA values in the internal capsule at 30 weeks’ gestational age.98 Krishnan and colleagues99 reported a significant negative correlation between ADC values in the centrum semiovale and the Griffiths mental scale score at the age of 2 years, and several other investigators have found highly significant correlations between FA values in both the posterior limb of the internal capsule100 and the splenium of the corpus callosum101 and neurodevelopmental outcome measures in preterm children who were scanned at term-equivalent age and examined in early childhood. 1047

Review

Correlations between DTI values and visual assessment scores have also been found.102 The FA values in the optic radiations increase with gestational age,103 and FA values from both tract-based spatial statistics analyses and probabilistic tractography indicate that visual function is directly associated with FA in the optic radiations of preterm infants at term-equivalent age.102,103

Functional connectivity Several resting state networks are reported to be present in preterm infants at term-equivalent age.64 Similar to term controls,63 preterm neonates had resting state networks that encompassed the primary visual cortex, bilateral sensorimotor areas, bilateral auditory cortex, a network including the precuneus, lateral parietal cortex and the cerebellum, and an anterior region that incorporated the medial and dorsolateral prefrontal cortex. Several preliminary reports have indicated the presence of multiple resting state networks in preterm infants as early as 30 weeks’ postmenstrual age. Whether these networks are intrinsic to the developing brain or mature across the third trimester remains an area of active investigation.104–106 The effect of preterm birth on

resting state functional connectivity networks is just beginning to be investigated.

Imaging the preterm brain: childhood and adolescence Although many progressive and regressive events that are crucial for cerebral development occur during the third trimester of gestation and in the newborn period, both postmortem examinations and MRI studies of typically developing subjects suggest that dynamic changes in brain composition and connectivity continue during childhood, adolescence, and early adulthood.39,107–109 In typically developing subjects, there are linear increases in white matter of 12–25% between the ages of 4 and 20 years, while synapses are pruned and grey matter decreases in size.110,111

Structure–function relationships DTI studies DTI abnormalities have been reported in prematurely born children during childhood and adolescence (table 3 and figure 3).112–115 Similar to neonatal data, these abnormalities appear to cluster in central white matter and the regions to which it connects. In 33 children who were born

Number

Birthweight/ Major exclusion criteria Age at MRI GA

Arzoumanian et al96

137 preterm

<1800 g

Congenital malformation; gross structural abnormalities on MRI

Als et al97

30 preterm

28–33 weeks

Congenital abnormalities 2 weeks’ CA and infections; abnormal MRIs

Early intervention study: FA Experimental group had increased relative anisotropy in left PLIC compared with control group (p=0·03) at 2 weeks; behaviour at Correlation between DTI and behaviour 9 months; ROI

Drobyshevsky 24 preterm et al98

<32 weeks

Congenital anomalies

Serial: 30 and 36 weeks’ GA

No injury, grade 1–2 IVH, and grade 3–4 IVH groups: ADC, FA, Bayley PDI at 18–24 months’ CA; ROI

Krishnan et al99

38 preterm

<33 weeks

Overt pathology on conventional MRI

Term

Significant negative correlation between mean ADC and developmental Griffiths mental scores developmental scale; ADC of centrum semiovale

Rose et al100

24 preterm

<1800 g

Congenital anomalies and infections; blindness; quadruplets

Term

GMFS, cerebral palsy; ROI

Rose et al101

78 very low GA 27·6±2·5 birthweight weeks, birthweight 1021±339 g

Congenital anomalies and infections

Bayley PDI scales and 37 weeks’ postmenstrual GMFS at 18 months’ CA; ROI age

Male infants had lower FA and higher ADC than did female infants (p<0·05) Abnormal development more common in males than females (p<0·05) Children with abnormal development had decreased FA in splenium and right PLIC (p<0·05) Logistic regression predicted development with 94% accuracy; only the right PLIC was a significant logistic component (p=0·015)

Bassi et al102

27 preterm

24–32 weeks

Congenital anomalies

2 weeks’ CA

Correlation between visual assessment score and FA value in optic radiations (p<0·001)

Berman et al103

36 preterm

21–49 weeks

NA

29–41 weeks’ DTI fibre tracking of optic postmenstrual radiations; visual fixation task; probabilistic fibre age tracking

Term

Outcome measures

Results

Neurological examination, cerebral palsy at 18–24 months; ROI

Children with cerebral palsy had lower FA of right PLIC than did healthy infants (p=0·0006) Children with abnormal examination had lower FA of right PLIC than did healthy infants (p=0·005)

Visual assessment score; TBSS, probabilitistic fibre tracking

Decreased FA and increased ADC in infants with grade 3–4 IVH at second MRI compared with preterm infants with no IVH Correlation between FA at 30 weeks’ GA and Bayley PDI, and inverse correlation between change in FA of internal capsule and occipital WM and Bayley PDI

Children with low neonatal FA had negative correlation between FA in left and right PLIC and normalcy index for severity of gait deficits (p=0·001; r=–0·89) and GMFS (p=0·04; r=–0·65)

FA in optic radiation increased with GA; mean diffusivity, parallel diffusivity, and transverse diffusivity decreased with GA Correlation between optic radiation FA and visual fixation tracking scores (p<0·006)

ADC=apparent diffusion coefficient. CA=corrected age. DTI=diffusion tensor imaging. FA=fractional anisotropy. GA=gestational age. GMFS=gross motor function scale. IVH=intraventricular haemorrhage. NA=not available. PDI=psychomotor developmental index. PLIC=posterior limb of internal capsule. ROI=region of interest. TBSS=tract-based spatial statistics. WM=white matter.

Table 2: Relations between WM alterations and developmental outcome in prematurely born subjects—DTI studies

1048

www.thelancet.com/neurology Vol 8 November 2009

Review

Number

Major exclusion criteria

Birthweight/ GA

Age at MRI

Outcome measures

Results

Nagy et al112

9 preterm with attention-deficit hyperactivity disorder, 10 term controls

IVH, PVL, full scale IQ <80

<1500 g

11 years

FA of WM; regions of interest

Preterm subjects had lower FA in posterior CC (p<0·001) and bilateral internal capsules (p<0·001) than did term controls

Counsell et al113

12 preterm

NA (minimal thinning of body of CC in 7, mild ventriculomegaly in 2, moderate ventriculomegaly in 1, porencephaly in 1)

Median 30 weeks’ postmenstrual age

Median 24·6 months’ corrected age (range 21–26·36 months)

Probabilistic tractography to visualise WM tracts and thalamocortical connections

In 11 preterm subjects with no focal lesions, investigators mapped cortical connections to the thalamus that appeared similar to those seen in adults In 1 child with WM abnormalities, volume and pattern of thalamocortical connections were severely disturbed

Yung et al114

25 normal preterm, 13 term

Congenital malformations and infections

<1500 g

8·8–11·5 years

FA of whole-brain WM; Preterm group had less WM volume (p=0·014) and FA than IQ did term group (p<0·001) Both volume (r²=0·407; p=0·021) and FA (r²=0·496; p=0·005) were independent predictors of full scale IQ

Skranes et al115

34 preterm, 47 term

NA (in the preterm group, 79% had ventriculomegaly, 44% had decreases in posterior WM, 41% had thinning of the posterior CC)

<1500 g

15 years

FA of WM; VMI, WISC-III; movement ABC, grooved pegboard, attention-deficit hyperactivity rating scale; regions of interest

Preterm subjects had reduced FA in internal and external capsule, CC, and superior, middle superior, and inferior fasciculus Adolescents born prematurely with low IQ had low FA in external capsule and inferior and middle superior fasciculus Visuomotor deficits correlated with low FA in external capsule, posterior limb of internal capsule, and inferior fasciculus (p<0·05 for all)

Counsell et al116

33 preterm

Median 28+5 Congenital abnormalities, metabolic disease, focal lesions or weeks’ GA posthaemorrhagic hydrocephalus on neonatal MRI (24 with diffuse excessive hyperintense signal, 14 ventriculomegaly, 6 germinal matrix haemorrhage; 1 IVH)

2 years’ corrected age

Neonatal DTI; Griffiths mental developmental scales; TBSS

Developmental quotient linearly associated with FA in isthmus and body of CC (p<0·01) Performance subscores associated with FA in CC (p<0·01) and right cingulum (p<0·01)

Constable et al117

29 preterm, 22 term

Congenital infections and malformations; IVH, PVL, and ventriculomegaly on neonatal ultrasound

600–1250 g

12 years

FA; WISC-III, VMI; tractography for uncinate

Differences in FA in right inferior fronto-occipital fasciculus, anterior uncinate fasciculi, deep WM, splenium (p<0·01 for all) FA in left anterior uncinate correlated with verbal IQ, full scale IQ, PPVT for preterm males (p<0·05)

Murakimi et al118

10 preterm with PVL

NA

24+4 to 34+3 weeks’ GA

19±9·5 months

Motor examination at 28±14·5 months; tractography

No difference in FA values for bilateral centrum semiovale or genus and splenium of CC between children with or without CP Children with CP had lower FA of motor tracts than did those without CP (p<0·001) FA <0·5 predicted severe CP (p<0·001; r²=1)

Kontis et al119

63 very preterm, 45 term (all right-handed)

NA

<33 weeks’ GA 19 years (range 17–22 years)

FA and MD of genu, body, and splenium of CC, tractography

Very preterm females had higher MD in the genu than did term females (p=0·004), which was associated with lower performance IQ For preterm group, MD of body of CC was significantly correlated with intrusions subtest of the CVLT; for term group, genu MD and splenium MD were also associated with the learning slope subscore (p<0·005 for both)

CC=corpus callosum. CP=cerebral palsy. CVLT=California verbal learning test. DTI=diffusion tensor imaging. FA=fractional anisotropy. GA=gestational age. IQ=intelligence quotient. IVH=intraventricular haemorrhage. MD=mean diffusivity. Movement ABC=movement assessment battery for children. NA=not available. PPVT=Peabody picture vocabulary test. PVL=periventricular leukomalacia. TBSS= tract-based spatial statistics. VMI=visuo-motor integration test. WISC=Wechsler intelligence scale. WM=white matter.

Table 3: Effects of preterm birth on WM connectivity in childhood and adolescence—DTI studies

prematurely at 2 years’ corrected age, the developmental quotient was linearly associated with FA in parts of the corpus callosum, whereas performance subscores were correlated with FA values in both the corpus callosum and the right cingulum.116 DTI parameters in prematurely born school-aged children, adolescents, and young adults have also been shown to significantly correlate with behavioural and motor outcomes.117–119 Nagy and colleagues112 noted alterations in FA values for both internal capsules and the corpus callosum in preterm males with attention-deficit www.thelancet.com/neurology Vol 8 November 2009

hyperactivity disorder and ten matched controls at 11 years’ corrected age. Vangberg and colleagues120 reported decreased FA values in several white matter regions, including the internal capsule, corpus callosum, and superior fasciculus, in preterm subjects compared with term controls at the age of 15 years. Preterm adolescents with low intelligence scores had low FA values in the external capsule and inferior and middle superior fasciculi; furthermore, visuomotor deficits correlated with low FA values in the posterior limb of the internal capsule, external capsule, and inferior fasciculus.115 1049

Review

B

A

150 140 130 120

PPVT–R

110 100 90 80 70 60 50 0·25

0·26

0·27

0·28

0·29

0·30

0·31

0·32

0·33

0·34

0·35

0·36

0·37

Left uncinate fractional anisotropy

Figure 3: Fractional anisotropy values in the left uncinate correlate with PPVT–R scores in preterm patients at 16 years of age Axial diffusion tensor image showing anisotropy in the left uncinate in 44 preterm patients at 16 years of age (A). Graph of fractional anisotropy values and PPVT-R scores for the patients (p=0·02; r²=0·32; B). Figure courtesy of Cheryl Lacadie (Yale University School of Medicine, New Haven, CT, USA). PPVT–R=Peabody picture vocabulary test–revised.

Similarly, Constable and colleagues117 reported widespread decreases in FA in preterm children compared with term controls at the age of 12 years. Changes were found in intrahemispheric association fibres subserving language skills (ie, the external capsule and the anterior portions of the uncinate fasciculi bilaterally) and the deep white matter regions to which they project, as well as the splenium of the corpus callosum. Finally, after studying prematurely born adolescents and term control subjects at 19 years of age, Kontis and colleagues119 found that the Dav in the body of the corpus callosum was significantly correlated with verbal memory skills in the preterm group.

VBM studies VBM has been extensively used to examine the effects of preterm birth on the developing brain at adolescence and in early adulthood (table 4 and figure 4).121,122 Gimenez and colleagues123,124 examined preterm and term control subjects at ages 12–18 years using both modulated (volume based) and unmodulated (density) analyses. The unmodulated data showed regions in the periventricular white matter and longitudinal fascicles that were significantly smaller in the preterm subjects than in term controls, whereas in the modulated analyses there were white matter decreases in regions distant from the ventricles at the origins and ends of the longitudinal fascicles. For many regions throughout the brain, the lower the prematurely born subjects’ gestational age or birthweight, the lower the white matter volume found. Isaacs and colleagues121 identified a region in the left parietal lobe where there was less grey matter in prematurely born children with calculation deficits than in those without this disability. In the same population, 1050

absolute IQ scores were associated with changes in areas in both the parietal and temporal lobes.122 Similarly, Gimenez and colleagues125 noted significant decreases in thalamic nuclei when prematurely born subjects were compared with term controls at 14 years of age; there were significant correlations between thalamic nuclei volumes and tests of verbal fluency for the preterm group. These findings confirm the potential for decreases in thalamic volumes to be used as a biomarker of injury in preterm infants. New techniques such as automatic labelled brain atlases combined with conventional image analysis will further refine the structural assessment of the effects of prematurity.129 VBM studies also indicate widespread decreases in white matter at adolescence.55 Soria-Pastor and colleagues126 reported significant decreases in white matter in 80% of adolescents who were born prematurely (<32 weeks’ gestational age) when compared with term controls at the age of 14 years. There were significant alterations in the centrum semiovale in 41% of the preterm subjects, while more than one-fifth had lower periventricular white matter volumes than term controls. Nosarti and colleagues56 have studied the largest cohort of prematurely born adolescents and term controls, matched by age and socioeconomic status. They noted widespread alterations in both grey and white matter volumes throughout the cerebral hemispheres (table 4). All areas where there were between-group white and grey matter differences were structurally associated, and subjects with ventriculomegaly had the most significant white matter changes. Several white matter regions, including those within the corpus callosum, were linearly associated with gestational age for the preterm group, and decreases in white matter volume in the middle www.thelancet.com/neurology Vol 8 November 2009

Review

Number

Major exclusion criteria

Birthweight/ Age at MRI VBM template, GA outcome measures

Results

Isaacs et al121

12 preterm with deficits in numerical operations, 12 preterm with deficits in mathematical reasoning, 12 preterm controls for each group

NA

<30 weeks

15 years

VBM: template from 20 normal children; IQ, WOND

Subjects who had deficits in numerical operations had less GM in left intraparietal sulcus (p<0·001) than those with no numerical operation deficits No differences for group who had deficits in numerical operations compared with control group

Isaacs et al122

82: 4 groups (verbal IQ and performance IQ; small and large declines)

Neuromotor or neurosensory impairment

<30 weeks

15 years

VBM: template from 20 normal children; IQ at 7 and 15 years

Negative correlation between performance IQ decline and GM in hippocampi (p=0·03) Volumetric data confirmed hippocampi were 4–5% smaller in group with large decline in performance IQ compared with those without change in IQ

Gimenez et al123

50 preterm, 50 term

Evidence of WM injury on T2 MRI

≤32 weeks

12–18 years Modulated and unmodulated VBM analyses; T1 SPM2 template with WMspecific template

Gimenez et al124

22 preterm, 22 term

Focal TBI, CP, neurological diagnosis, MRI abnormalities

<32 weeks

14–15 years VBM: SPM2 normalised In preterm subjects, VBM indicated reduced GM volume in the orbital region compared with term controls brain; sulcal Preterm subjects had significant decrease in the secondary sulci depth but not in measurements primary sulci depth compared with term controls

Gimenez et al125

30 preterm, 30 term

Focal TBI, CP, neurological diagnosis, MRI abnormalities

<32 weeks

14 years

VBM: T1 SPM template; Preterm subjects had worse scores for both phonetic (p=0·044) and semantic phonetic and semantic fluency (p<0·001) than did term controls Preterm subjects had lower thalamic volumes than did term controls (left: fluency tasks p=0·001; right: p=0·002) Significant correlations between thalamic nuclei volume changes and verbal fluency (p<0·05) Semantic fluency was more highly correlated with thalamic nuclei GM than was phonetic fluency

Kesler et al55

29 preterm with no brain injury on ultrasound, 22 term

600–1250 g Congenital infection or anomalies; IVH, PVL, or ventriculomegaly on neonatal ultrasound

12 years

VBM: customised template from all study subjects; complete paediatric neuropsychological battery

Preterm males had lower WM in bilateral cingulum, CC, corticospinal tracts, prefrontal cortex, and superior and inferior longitudinal fasciculus than did term subjects (p<0·001); no difference for females Preterm males had less GM volume in prefrontal cortex, basal ganglia, and temporal lobe (p<0·001) than did term males; no difference for females No correlation between VBM and perinatal or cognitive data

Soria-Pastor et al126

44 preterm, 43 term

TBI, CP, seizures, MRI abnormalities

14 years

VBM: T1 SPM2 template; full scale IQ, verbal IQ, performance IQ and digit symbol subtest

Reduced global WM volume in preterm compared with term subjects (p<0·05) Performance IQ correlated with whole-brain WM volume (r=0·34; p=0·026) 80% preterm subjects had significant WM abnormalities; 41% had abnormal centrum semiovale; and 21% had abnormal periventricular WM Low scores on digit symbol subtest correlated with decreases in WM

Nosarti et al56

218 preterm, 128 term

<33 weeks Neurological diagnosis, meningitis, TBI, cerebral infection

Narberhaus et al127

21 preterm, 22 term

History of CP, grade 3–4 IVH, PVL

Nosarti et al128

28 preterm, 26 term

Neurological impairment

<32 weeks

Unmodulated data identified regions in the periventricular WM and longitudinal fascicles that had significantly lower volume in preterm compared with term controls Modulated data showed WM decreases in regions distant from the ventricles at the origin and end of the longitudinal fascicles For several regions throughout the brain, the lower the GA and/or birthweight, the lower the WM volume

14–15 years VBM of GM and WM: customised template from all study subjects; measures of reading, spelling, IQ, executive function, visuomotor integration test

Widespread alterations in GM and WM throughout brain for preterm subjects All areas where between-group WM and GM differences were observed were structurally associated Preterm subjects with IVH and ventriculomegaly had greatest WM and GM alterations Preterm subjects had lower scores on language and executive function and were more likely to show cognitive impairment compared with term controls Several GM (sensorimotor, middle, superior temporal gyri) and WM (precentral and postcentral gyri, longitudinal fasciculus, CC) areas were linearly associated with GA; middle temporal gyrus (both WM and GM) associated with cognitive impairment in preterm group (p<0·001)

<33 weeks

20 years

Preterm and term subjects had differential BOLD signal during both encoding and recognition Preterm subjects had decreased hippocampal GM compared with term controls

<33 weeks

20–21 years fMRI, VBM; phonological verbal fluency fMRI task

fMRI: visual paired associates task; VBM of hippocampus

Significant differences in BOLD signal for both “easy” and “hard” words between the groups Altered GM distribution in the preterm infants compared with control infants fMRI results only partly explained by volumetric differences

BOLD=blood oxygen level dependent. CC=corpus callosum. CP=cerebral palsy. fMRI=functional MRI. GA=gestational age. GM=grey matter. IQ=intelligence quotient. IVH=intraventricular haemorrhage. NA=not available. PVL=periventricular leukomalacia. SPM=statistical parametric mapping. TBI=traumatic brain injury. VBM=voxel-based morphometry. WM=white matter. WOND=Wechsler objective numeric dimension test.

Table 4: Relations between GM and WM alterations and outcome in prematurely born children and adolescents—VBM studies

www.thelancet.com/neurology Vol 8 November 2009

1051

Review

associates test in young adults born prematurely. Despite good task performance, preterm subjects had activation of different neural networks during mnemonic processing of visuo-perceptual material; additionally, the preterm group had decreased hippocampal grey matter bilaterally. Nosarti and colleagues128 used the same strategies on preterm subjects and term controls to detect alterations in brain activation during completion of a letter fluency task and compared this with regional VBM data. BOLD signals differed significantly between groups, but fMRI results were only partly explained by structural changes.

Functional connectivity Figure 4: Optimised voxel-based morphometry group by sex effect (ie, term males minus preterm males) Decreased grey matter volume was seen in the left inferior frontal gyrus, or Broca’s region, at 12 years of age in preterm males (p<0·001; data from Kesler and colleagues55). Figure courtesy of Shelli Kesler (Stanford University School of Medicine, Palo Alto, CA, USA). Effect of gestational age

Effect of injury

Regions of interest: Regions linearly associated with gestational grey matter age include the middle and superior temporal gyri, inferior frontal gyrus, medial frontal gyrus, fusiform gyrus, and thalamus

Decreased volume in the thalamus and pulvinar Increased volume in the medial frontal gyrus and precuneus

Regions of interest: Regions linearly associated with gestational white matter age include the middle temporal gyrus, medial frontal gyrus, fusiform gyrus, anterior corpus callosum, parahippocampal gyrus, and supramarginal gyrus

Decreased volume in the middle temporal gyrus, cingulum, and posterior corpus callosum Increased volume in the fusiform, parietal lobe, and middle occipital gyrus

Table 5: Effects of preterm birth on grey and white matter in adolescence—voxel-based morphometry studies

Search strategy and selection criteria References for this Review were identified through searches of PubMed for MRI studies in prematurely born individuals published between January, 1985, and August, 2009. We used combinations of the following search terms: “infant”, “neonate”, “child”, “adolescent”, “adult”, “preterm”, “premature”, “prematurely born”, “brain”, “magnetic resonance imaging”, “MRI”, “functional magnetic resonance imaging”, “fMRI”, “functional connectivity”, “fcMRI,” “resting state connectivity”, “diffusion tensor imaging”, “DTI”, “tractography”, “probabilistic tractography”, “TBSS (tract-based spatial statistics)”, “volumetric”, “voxel-based morphometry”, “VBM”, “cortical folding”, “gyrification”, and “cortical surface area”. We reviewed the published studies and selected those that we judged to have most relevance to the topic. Only papers published in English were included. Preclinical studies, case reports, unblinded analyses, and topic reviews were excluded.

temporal gyrus were associated with cognitive impairment in the preterm group (table 5). Finally, studies in typically developing children have documented the correlation between structure and function.130,131 Narberhaus and colleagues127 investigated the neural substrates of visual memory using a visual paired 1052

Over time, the connections in spatially remote regions in the developing brain are strengthened in an anterior to posterior direction, weaving distal brain areas into highly cohesive and connected circuits, while the strength of connections between contralateral homologues is reduced.66,67 When Gozzo and colleagues132 used a passive language task to study fcMRI in preterm children and term controls at the age of 8 years, those born prematurely had significantly stronger neural circuits between Wernicke’s area and alternative language regions including the right inferior frontal gyrus and both left and right supramarginal gyri. These data suggest that either prematurely born children use alternative networks for neural processing of language, or that there is a delay in maturation in the preterm group; additional studies are needed to verify this hypothesis.

Conclusions and future perspectives Preterm birth is a common event, and injury to the developing brain occurs far too often in the prematurely born. Imaging technologies that enable exploration of acute injury and post-lesional alterations in structural and functional connectivity associated with preterm birth are crucially important—not only for assessing long-term neurodevelopmental risk but also for planning both individualised treatment and therapeutic intervention trials. The outcome of future neuroprotective treatments ranging from pharmacological drugs to stem cell therapies and environmental enrichment will require the routine monitoring of anatomic integrity, microstructural connectivity, and functional performance. DTI-based imaging technologies are being developed to better examine the effect of preterm birth on axonal development. Future work is needed to investigate both strategies for automatic standardised segmentation of the preterm brain as well as the generation of a neonatal neuroimaging atlas that would provide uniform nomenclature among investigators. Longitudinal studies will provide important information about the dynamic effect of preterm birth at adolescence and beyond, and the refinement of a combined structure–function approach, such as use of the recently developed technique of diffusion spectrum imaging with the establishment of high-resolution connection matrix and resting state fcMRI, www.thelancet.com/neurology Vol 8 November 2009

Review

will inform both physicians and scientists about the processes of ongoing plasticity in vivo. Preterm birth is now a major paediatric public health problem. If the goal of neonatal intensive care is survival without handicap for prematurely born infants, then identifying those early biomarkers of injury and outcome is crucial for our understanding of the developing brain. If connectivity can be altered and plasticity understood, targeted pharmacological and behavioural therapies could be developed to restore functional connectivity.

15

16

17

18

Contributors All authors contributed equally to this paper.

19

Conflicts of interest We have no conflicts of interest.

20

Acknowledgments This work was supported by grants NS 27116 and NS 53865 from the National Institute of Neurological Disorders and Stroke, National Institutes of Health (LRM), the Swiss National Science Foundation SNF 32-102127; 32-113632 (PSH), the Leenards Foundation, and the European Union 6th Framework Programme NEOBRAIN (PSH). The authors are grateful to Cheryl Lacadie, Karol H Katz, and Jill Maller-Kesselman for technical, computer, and editorial support and to the children and families who participated in the studies we describe. References 1 Behrman RE, Stith Butler A. Institute of Medicine Committee on understanding premature birth and assuring healthy outcomes board on health sciences outcomes: preterm birth: causes, consequences, and prevention. Washington, DC, USA: the National Academies Press, 2007. 2 Allen MC. Neurodevelopmental outcomes of preterm infants. Curr Opin Neurol 2008; 21: 123–28. 3 Robertson CM, Howarth TM, Bork DL, Dinu IA. Permanent bilateral sensory and neural hearing loss of children after neonatal intensive care because of extreme prematurity: a thirty-year study. Pediatrics 2009; 123: e797–807. 4 Hack M, Taylor HG, Schluchter M, Andreias L, Drotar D, Klein N. Behavioral outcomes of extremely low birth weight children at age 8 years. J Dev Behav Pediatr 2009; 30: 122–30. 5 Rushing S, Ment LR. Preterm birth: a cost benefit analysis. Semin Perinatol 2004; 28: 444–50. 6 Latal B. Prediction of neurodevelopmental outcome after preterm birth. Ped Neurol 2009; 40: 413–19. 7 Mikkola K, Ritari N, Tommiska V, Salokorpi T, Lehtonen L, Tammela O. Neurodevelopmental outcome at 5 years of age of a national cohort of extremely low birth weight infants who were born in 1996–1997. Pediatrics 2005; 116: 1391–400. 8 Larroque B, Ancel P-Y, Marret S, et al. Neurodevelopmental disabilities and special care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): a longitudinal cohort study. Lancet 2008; 371: 813–20. 9 Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet 2008; 371: 261–69. 10 Fanaroff AA, Stoll BJ, Wright LL, et al. Trends in neonatal morbidity and mortality for very low birthweight infants. Am J Ostet Gynecol 2007; 196: 147.e1–8. 11 Marlow N, Wolke D, Bracewell MA, Samara M. Neurologic and developmental disability at six years of age after extremely preterm birth. N Engl J Med 2005; 352: 9–19. 12 Hüppi PS, Schuknecht B, Boech C, Bossi E. Structural and neurobehavioral delay in postnatal brain development of preterm infants. Pediatr Res 1996; 39: 895–901. 13 Mercuri E, Jongmans M, Henderson S, et al. Evaluation of the corpus callosum in clumsy children born prematurely: a functional and morphological study. Neuropediatrics 1996; 27: 317–22. 14 de Vries CJ, Groenendaal F, van Haastert I-LC, Eken P, Rademaker KJ, Meiners LC. Asymmetrical myelination of the posterior limb of the internal capsule in infants with periventricular haemorrhagic infarction: an early predictor of hemiplegia. Neuropediatrics 1999; 30: 314–19.

www.thelancet.com/neurology Vol 8 November 2009

21

22

23

24

25

26 27

28

29

30 31

32

33

34

35

36 37 38

39

Maalouf EF, Duggan PJ, Rutherford MA, et al. Magnetic resonance imaging of the brain in a cohort of extremely preterm infants. J Pediatr 1999; 135: 351–57. Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal MRI to predict neurodevelopmental outcomes in preterm infants. N Engl J Med 2006; 355: 685–94. Hagmann CF, De Vita E, Bainbridge A, et al. T2 at MR imaging is an objective quantitative measure of cerebral white matter signal intensity abnormality in preterm infants at term-equivalent age. Radiology 2009; 252: 209–17. Hüppi PS, Warfield S, Kikinis R, et al. Quantitative magnetic resonance imaging of brain development in premature and mature newborns. Ann Neurol 1998; 43: 224–35. Boardman JP, Counsell SJ, Rueckert D, et al. Early growth in brain volume is preserved in the majority of preterm infants. Ann Neurol 2007; 62: 185–92. Peterson BS, Vohr B, Cannistraci CJ, et al. Regional brain volume abnormalities and long-term cognitive outcome in preterm infants. JAMA 2000; 284: 1939–47. Allin M, Henderson M, Suckling J, et al. Effects of very low birthweight on brain structure in adulthood. Dev Med Child Neurol 2004; 46: 46–53. Inder TE, Hüppi PS, Warfield S, et al. Periventricular white matter injury in the premature neonate is followed by reduced cerebral cortical gray matter volume at term. Ann Neurol 1999; 46: 755–60. Inder TE, Warfield S, Wang H, Hüppi PS, Volpe JJ. Abnormal cerebral structure is present at term in premature infants. Pediatrics 2005; 115: 286–94. Mewes AUJ, Hüppi PS, Als H, et al. Regional brain development in serial magnetic resonance imaging of low-risk preterm infants. Pediatrics 2006; 118: 23–33. Volpe JJ. Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. Lancet Neurol 2009; 8: 110–24. Hilgetag CC, Barbas H. Developmental mechanis of the primate cerebral cortex. Anat Embryol 2005; 210: 411–17. van Essen DC. A tension-based theory of morphogenesis and compact wiring in the central nervous system. Nature 1997; 385: 313–18. Neubauer A-P, Voss W, Kattner E. Outcome of extremely low birth weight survivors at school age: the influence of perinatal parameters on neurodevelopment. Eur J Pediatr 2008; 167: 87–95. Voss W, Neubauer A-P, Wachtendorf M, Verhey JF, Kattner E. Neurodevelopmental outcome in extremely low birth weight infants: what is the minimum age for reliable developmental prognosis? Acta Pediatr 2007; 96: 342–47. Aylward GP. Neurodevelopmental outcomes of infants born prematurely. Devel Behav Ped 2005; 26: 427–40. Bhutta AD, Cleves MA, Casey PH, Cradock MM, Anand KJS. Cognitive and behavioral outcomes of school-aged children who were born preterm. A meta-analysis. JAMA 2002; 288: 728–37. Rushe TM, Rifkin L, Stewart AL, et al. Neuropsychological outcome at adolescence of very preterm birth and its relation to brain structure. Devel Med Child Neurol 2001; 43: 226–33. Doyle LW, Casalaz D. Outcome at 14 years of extremely low birthweight infants: a regional study. Arch Dis Child Fetal Neonatal Ed 2001; 85: F159–64. Walther FJ, den Ouden AL, Veerloove-Vanhorick SP. Looking back in time: outcome of a national cohort of very preterm infants born in the Netherlands in 1983. Early Hum Devel 2000; 59: 175–91. Verrips E, Vogels T, Saigal S, et al. Health-related quality of life for extremely low birth weight adolescents in Canada, Germany and the Netherlands. Pediatrics 2008; 122: 556–61. Hack M, Klein N. Young adult attainments of preterm infants. JAMA 2006; 295: 695–96. Banker BQ, Larroche JC. Periventricular leukomalacia of infancy: a form of neonatal anoxic encephalopathy. Arch Neurol 1962; 7: 32–50. Haynes RL, Billiards SS, Borenstein NS, Volpe JJ, Kinney HC. Diffuse axonal injury in periventricular leukomalacia as determined by apoptotic marcter fractin. Pediatr Res 2008; 63: 656–61. Kostovic I, Jovanov-Milosevic N. The development of cerebral connections during the first 20–45 weeks’ gestation. Sem Fet Neo Med 2006; 11: 415–22.

1053

Review

40

41

42

43

44

45

46 47

48

49

50

51

52

53

54 55

56

57

58 59

60

61

62 63

64

1054

Kostovic I, Judas M, Rados M, Hrabac P. Laminar organization of the human fetal cerebrum revealed by histochemical markers and magnetic resonance imaging. Cerebral Cortex 2002; 12: 536–44. Dyet LE, Kennea NL, Counsell SJ, et al. Natural history of brain lesions in extremely preterm infants studied with serial magnetic resonance imaging from birth and neurodevelopmental assessment. Pediatrics 2006; 118: 536–48. Nosarti C, Al-Asady MH, Frangou S, Stewart A, Rifkin L, Murray RM. Adolescents who were born very preterm have decreased brain volumes. Brain 2002; 125: 1616–23. Lodygensky GA, Rademaker KJ, Zimine S, et al. Structural and functional brain development after hydrocortisone treatment for neonatal chronic lung disease. Pediatrics 2005; 116: 1–7. Boardman JP, Counsell SJ, Rueckert D, et al. Abnormal deep grey matter development following preterm birth detected using deformation-based morphometry. NeuroImage 2006; 32: 70–78. Weiss J, Takizawa B, McGee A, et al. Neonatal hypoxia suppresses oligodendrocyte Nogo-A and increases axonal sprouting in a rodent model for human prematurity. Exp Neurol 2004; 189: 141–49. Back SA, Riddle A, McClure MM. Maturation-dependent vulnerability of perinatal white matter in premature birth. Stroke 2007; 38: 724–30. Back SA. Perinatal white matter injury: the changing spectrum of pathology and emerging insightsinto pathogenetic mechanisms. Ment Retard Dev Disabil Res Rev 2006; 12: 129–40. Riddle A, Luo NL, Manese M, et al. Spatial heterogeneity in oligodendrocyte lineage maturation and not cerebral blood flow predicts fetal ovine periventricular white matter injury. J Neurosci 2006; 26: 3045–55. Drobyshevsky A, Derrick M, Wyrwicz AM, et al. White matter injury correlates with hypertonia in an animal model of cerebral palsy. J Cereb Blood Flow Metab 2007; 27: 270–81. Segovia KN, McClure M, Maravec M, et al. Arrested oligodendrocyte lineage maturation in chronic perinatal white matter. Ann Neurol 2008; 63: 520–30. McClure MM, Riddle A, Manese M, et al. Cerebral blood flow heterogeneity in preterm sheep: lack of physiologic support for vascular boundary zones in fetal cerebral white matter. J Cereb Blood Flow Metab 2008; 28: 995–1008. Mukherjee P, McKinstry RC. Diffusion tensor imaging and tractography of human brain development. Neuroimag Clin N Am 2006; 16: 19–43. Mori S, Crain BJ, Chacko VP, van Zijl PCM. Three-dimensional tracking of axonal projects in the brain by magnetic resonance imaging. Ann Neurol 1999; 45: 265–69. Ashburner J, Friston KJ. Why voxel-based morphometry should be used. NeuroImage 2001; 14: 1238–43. Kesler SR, Reiss AL, Vohr B, et al. Brain volume reductions within multiple cognitive systems in male preterm children at age twelve. J Pediatr 2008; 152: 513–20. Nosarti C, Giouroukou E, Healy E, et al. Grey and white matter distribution in very preterm adolescents mediates neurodevelopmental outcome. Brain 2008; 181: 205–17. Smith SM, Jenkinson M, Jonsen-Berg H, et al. Tract-based spatial sstatistics: voxelwise analysis of multi-subject diffusion data. NeuroImage 2006; 31: 1487–505. Hüppi PS, Dubois J. Diffusion tensor imaging of brain development. Semin Fetal Neonatal Med 2006; 11: 489–97. Counsell SJ, Boardman JP. Differential brain growth in the infant born preterm: current knowledge and future developments from brain imaging. Semin Fetal Neonatal Med 2005; 10: 403–10. Assaf Y, Blumenfeld-Katzir T, Yovel Y, Basser PJ. AxCaliber: a method for measuring axon diameter distribution from diffusion MRI. Magn Reson Med 2008; 59: 1347–54. Hagmann P, Kurant M, Gigandet X, et al. Mapping human whole-brain structural networks with diffusion MRI. PLoS One 2007; 2: e597. Hagmann P, Cammoun L, Gigandet X, et al. Mapping the structural core of human cerebral cortex. PLoS Biology 2008; 6: 1479–93. Fransson P, Skiold B, Engstrom M, et al. Spontaneous brain activity in the newborn brain during natural sleep—an fMRI study in infants born at full term. Pediatr Res 2009; 66: 301–05. Fransson P, Skiold B, Horsch S, et al. Resting-state networks in the infant brain. PNAS 2007; 104: 15531–36.

65

66 67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

Schafer RJ, Lacadie C, Vohr B, et al. Alterations in functional connectivity for language in prematurely-born adolescents. Brain 2009; 132: 661–70. Fair DA, Cohen AL, Dosenbach NUF, et al. The maturing architecture of the brain’s default network. PNAS 2008; 105: 4028–32. Fair DA, Dosenbach NUF, Church JA, et al. Development of distinct control networks through segregation and integration. PNAS 2007; 104: 13507–12. Greicius MD, Supekar K, Menon V, Dougherty RF. Resting-state functional connectivity reflects structural connectivity in the default mode network. Cereb Cortex 2009; 19: 72–78. Vincent JL, Patel GH, Fox MD, Snyder AZ, Baker JT, Van Essen DC. Intrinsic functional architecture in the anaethetized monkey brain. Nature 2007; 447: 83–86. Hampson M, Tokogl F, Sun Z, et al. Connectivity–behavior analysis reveals that functional connectivity between left BA 39 and Broca’s area varies with reading ability. NeuroImage 2006; 31: 513–19. Ment LR, Constable RT. Injury and recovery in the developing brain: evidence from functional MRI studies of prematurely-born children. Nat Clin Prac Neurol 2007; 3: 558–71. Dubois J, Benders M, Borradori-Tolsa C, et al. Primary cortical folding in the human newborn: an early marker of later functional development. Brain 2008; 131: 2028–41. Dubois J, Benders M, Cachia A, et al. Mapping the early cortical folding process in the preterm newborn brain. Cereb Cortex 2008; 18: 1444–54. Cachia A, Manqin JF, Riviere D, et al. A generic framework for the parcellation of the cortical surface into gyri using geodesic Voronoi diagrams. Med Image Anal 2003; 7: 403–16. Hüppi PS, Maier SE, Peled S, et al. Microstructural development of human newborn cerebral white matter assessed in vivo by diffusion tensor magnetic resonance imaging. Pediatr Res 1998; 44: 584–90. Hüppi PS, Murphy B, Jaier SE, et al. Microstructural brain development after perinatal cerebral white matter injury assessed by diffusion tensor magnetic resonance imaging. Pediatrics 2001; 107: 455–60. Miller SP, Vigneron DB, Henry RG, Bohland MA, Ceppi-Cozzio C, Hoffman C. Serial quantitative diffusion tensor MRI of the premature brain: development in new-borns with and without injury. J Magn Reson Imaging 2002; 16: 621–32. Partridge SC, Mukherjee P, Henry RG, et al. Diffusion tensor imaging: serial quantitation of white matter tract maturity in premature newborns. NeuroImage 2004; 22: 1302–14. Berman JI, Mukherjee P, Partridge SC, et al. Quantitative diffusion tensor MRI fiber tractography of sensorimotor white matter development in premature infants. NeuroImage 2005; 27: 862–71. Anjari M, Srinivasan L, Allsop JM, et al. Diffusion tensor imaging with tract-based spatial statistics reveals local white matter abnormalities in preterm infants. NeuroImage 2007; 35: 1021–27. Dudink J, Lequin M, van Pul C, et al. Fractional anisotropy in white matter tracts of very-low-birth-weight infants. Pediatr Radiol 2007; 37: 1216–23. Gimenez M, Miranda MJ, Born AP, Nagy Z, Rostrup E, Jernigan TL. Accelerated cerebral white matter development in preterm infants: a voxel-based morphometry study with diffusion tensor MR imaging. NeuroImage 2008; 41: 728–34. Rose SE, Hatzigeorgiou X, Strudwick MW, Durbridge G, Davies PSW, Colditz PB. Altered white matter diffusion anisotropy in normal and preterm infants at term-equivalent age. Magn Reson Med 2008; 60: 761–67. Cheong JLY, Thompson DK, Wang HX, et al. Abnormal white matter signal on MR imaging is related to abnormal tissue microstructure. AJNR Am J Neuroradiol 2009; 30: 623–28. Anjari M, Counsell SJ, Srinivasan L, et al. The association of lung disease with cerebral white matter abnormalities in preterm infants. Pediatrics 2009; 124: 268–76. Neil J, Miller J, Mukherjee P, Hüppi PS. Diffusion tensor imaging of normal and injured developing human brain—a technical review. NMR Biomed 2002; 15: 543–52. Srinivasan L, Dutta R, Counsell S, et al. Quantification of deep gray matter in preterm infants at term-equivalent age using manual volumetry of 3-tesla magnetic resonance images. Pediatrics 2007; 119: 759–65.

www.thelancet.com/neurology Vol 8 November 2009

Review

88

89

90

91 92 93

94

95

96

97 98

99

100

101

102

103

104

105

106

107 108

109

Ramenghi LA, Fumagalli M, Righini A, et al. Magnetic resonance imaging assessment of brain maturation in preterm neonates with punctate white matter lesions. Neuroradiology 2007; 49: 161–67. Battin MR, Maalouf EF, Counsell SJ, et al. Magnetic resonance imaging of the brain in very preterm infants: visualization of the germinal matrix, early myelination, and cortical folding. Pediatrics 1998; 101: 957–62. Childs A-M, Ramenghi LA, Cornette L, et al. Cerebral maturation in premature infants: Quantitative assessment using MR imaging. AJNR Am J Neuroradiol 2001; 22: 1577–82. Bystron I, Blakemore C, Rakic P. Development of the human cerebral cortex: Boulder Committee revisited. Nat Rev Neurosci 2008; 9: 110–22. Toro R, Burnod Y. A morphogenetic model for the development of cortical convolutions. Cereb Cortex 2005; 15: 1900–13. Biagioni E, Frisone MF, Laroche S, et al. Maturation of cerebral electrical activity and development of cortical folding in young very preterm infants. Clin Neurophys 2007; 118: 53–59. Vasileiadis GT, Thompson RT, Han VKM, Gelman N. Females follow a more “compact” early human brain development model than males. A case-control study of preterm neonates. Pediatr Res 2009; published online Aug 5, 2009. DOI:10.1203/PDR.0b013e3181ba1ae7. Inder TE, Hüppi PS, Zientara GP, et al. The postmigrational development of polymicrogyria documented by magnetic resonance imaging from 31 weeks’ postconceptional age. Ann Neurol 1999; 45: 798–801. Arzoumanian Y, Mirmiran M, Barnes PD, et al. Diffusion tensor brain imaging findings at term-equivalent age may predict neurologic abnormalities in low birth weight preterm infants. AJNR Am J Neurorad 2003; 24: 1646–53. Als H, Duffy FH, McAnulty GB, et al. Early experience alters brain function and structure. Pediatrics 2004; 114: 1738–39. Drobyshevsky A, Bregman J, Storey P, et al. Serial diffusion tensor imaging detects white matter changes that correlate with motor outcome in premature infants. Dev Neurosci 2007; 29: 289–301. Krishnan ML, Dyet LE, Boardman JP, et al. Relationship between white matter apparent diffusion coefficients in preterm infants at term-equivalent age and developmental outcome at 2 years. Pediatrics 2007; 120: e604–09. Rose J, Mirmiran J, Butler EE, et al. Neonatal microstructural development of the internal capsule on diffusion tensor imaging correlations with severity of gait and motor deficits. Devel Med Child Neurol 2007; 49: 745–50. Rose J, Butler EE, Lamont LE, Barnes PD, Atlas SW, Stevenson DK. Neonatal brain structure on MRI and diffusion tensor imaging, sex and neurodevelopment in very-low-birthweight preterm children. Devel Med Child Neurol 2009; 51: 526–35. Bassi L, Ricci D, Volzone A, et al. Probabilistic diffusion tractography of the optic radiations and visual function in preterm infants at term equivalent age. Brain 2008; 131: 573–82. Berman JI, Glass HC, Miller SP, et al. Quantitative fiber tracking analysis of the optic radiation correlated with visual performance in premature newborns. AJNR Am J Neuroradiol 2009; 30: 120–24. Doria V, Beckmann CF, Rees G, et al. Spontaneous brain activity in infants from 33 weeks gestational age. Society for Neuroscience; Washington, DC, USA; Nov 15–19, 2008. Program No. 787.13. Poster No. SS50. Doria V, Beckmann CF, Rees G, Counsell SJ, Merchant N, Edwards DA. Spontaneous activity in the developing preterm brain: a functional magnetic responance imaging study. PAS Abstract Book 2009: 2525.1. http://www.abstracts2view.com/pas/view. php?nu=PAS09L1_1423 (accessed Sept 18, 2009). Smyser CD, Inder TE, Hill JE, et al. Longitudinal analysis of neural network development in preterm infants. PAS Abstract Book 2009: 4765.4. http://www.abstracts2view.com/pas/view.php?nu=PAS09L1_ 1658 (accessed Sept 18, 2009). Toga AW, Thompson PM, Sowell ER. Mapping brain maturation. Trends Neurosci 2006; 29: 148–59. Sowell ER, Thompson PM, Leonard CM, Welcome SE, Kan E, Toga AW. Longitudinal mapping of cortical thickness and brain growth in normal children. J Neurosci 2004; 24: 8223–31. Yakovlev PI, Lecours AR. The myelogenetic cycles of regional maturation of the brain. In: Minkowski A, ed. Regional Development of the Brain in Early Life. Oxford: Blackwell Scientific; 1967: 3–70.

www.thelancet.com/neurology Vol 8 November 2009

110 Giedd JN, Lalonde FM, Celano MJ, et al. Anatomical brain magnetic resonance imaging of typically developing children and adolescents. J Am Acad Child Adolesc Psych 2009; 48: 465–70. 111 Sowell ER, Peterson BS, Thompson PM, Welcome SE, Henkenius AL, Toga AW. Mapping cortical change across the human life span. Nat Neurosci 2003; 6: 309–15. 112 Nagy Z, Westerberg H, Skare S, et al. Preterm children have disturbances of white matter at 11 years of age as shown by diffusion tensor imaging. Pediatr Res 2003; 54: 672–79. 113 Counsell SJ, Dyet LE, Larkman DJ, et al. Thalamo-cortical connectivity in children born preterm mapped using probabillistic magnetic resonance tractography. NeuroImage 2007; 34: 896–904. 114 Yung A, Poon G, Qui D, et al. White matter volume and anisotropy in preterm children: a pilot study of neurocognitive correlates. Ped Res 2007; 61: 732–36. 115 Skranes JS, Vangberg TR, Kulseng S, et al. Clinical findings and white matter abnormalities seen on diffusion tensor imaging in adolescents with very low birth weight. Brain 2007; 130: 654–66. 116 Counsell SJ, Edwards AD, Chew ATM, et al. Specific relations between neurodevelopmental abilities and white matter microstructure in children born preterm. Brain 2008; 131: 3201–08. 117 Constable RT, Ment LR, Vohr BR, et al. Prematurely born children demonstrate white matter microstructural differences at age 12 years relative to term controls: an investigation of group and gender effects. Pediatrics 2008; 121: 306–16. 118 Murakami A, Morimoto M, Yamada K, et al. Fiber-tracking techniques can predict the degree of neurologic impairment for periventricular leukomalacia. Pediatrics 2008; 122: 500–06. 119 Kontis D, Catani M, Cuddy M, et al. Diffusion tensor MRI of the corpus callosum and cognitive function in adults born preterm. NeuroReport 2009; 20: 424–28. 120 Vangberg TR, Skranes JS, Dale AM, Martinussen M, Brubakk A-M, Haraldseth O. Changes in white matter diffusion anisotropy in adolescents born prematurely. NeuroImage 2006; 32: 1538–48. 121 Isaacs EB, Edmonds C, Lucas A, Gadian DS. Calculation difficulties in children of very low birthweight: a neural correlate. Brain 2001; 124: 60–66. 122 Isaacs EB, Edmonds CJ, Chong WK, Lucas A, Morley R, Gadian DG. Brain morphometry and IQ measurement in preterm children. Brain 2004; 127: 2595–607. 123 Gimenez M, Junque C, Narberhaus A, Bargallo N, Botet F, Mercader JM. White matter volume and concentration reductions in adolescents with history of very preterm birth: a voxel-based morphometry study. NeuroImage 2006; 32: 1485–98. 124 Gimenez M, Junque C, Vendrell P, et al. Abnormal orbitofrontal development due to prematurity. Neurology 2006; 67: 1818–22. 125 Gimenez M, Junque C, Narberhaus A, Botet F, Bargallo N, Mercader JM. Correlations of thalamic reductions with verbal fluency impairment in those born prematurely. Neuroreport 2006; 17: 463–66. 126 Soria-Pastor S, Gimenez M, Narberhaus A, et al. Patterns of cerebral white matter damage and cognitive impairment in adolescens born very preterm. Int J Devel Neuroscience 2008; 26: 647–54. 127 Narberhaus A, Lawrence E, Allin MP, et al. Neural substrates of visual paired associates in young adults with a history of very preterm birth: alterations in fronto-parieto-occipital networks and caudate nucleus. NeuroImage 2009; 47: 1884–93. 128 Nosarti C, Sukhwinder S, Shergill S, et al. Neural substrates of letter fluency processing in young adults who were born very preterm: alterations in frontal and striatal regions. NeuroImage 2009; 47: 1904–13. 129 Gousias IS, Rueckert D, Heckemann RA, et al. Automatic segmentation of brain MRIs of 2-year-olds into 83 regions of interest. NeuroImage 2008; 40: 672–84. 130 Honey CJ, Sporns O, Cammoun L, et al. Predicting human resting-state functional connectivity from structural connectivity. PNAS 2009; 106: 2035–40. 131 Skudlarski P, Jagannathan K, Calhoun VD, Hampson M, Skudlarski BA, Pearlson G. Measuring brain connectivity: Diffusion tensor imaging validates resting state temporal correlations. NeuroImage 2008; 43: 445–61. 132 Gozzo Y, Vohr B, Lacadie C, et al. Alterations in neural connectivity in preterm children at school age. NeuroImage 2009; 48: 458–63.

1055