Abstract

Premature exposure to the extrauterine environment negatively affects the brains' developmental trajectory. Our aim was to determine whether extremely preterm (EPT) infants, with no evidence of focal brain lesions, show morphological brain differences when compared with term-born infants. Additionally, we investigated associations between perinatal factors and neuroanatomical alterations. Conventional magnetic resonance imaging was acquired at term-equivalent age (TEA) from 47 EPT infants born before 27 weeks of gestation, and 15 healthy, term-born controls. Automatic segmentation and voxel-based morphometry-Diffeomorphic Anatomical Registration through Exponentiated Lie algebra (DARTEL) were used. Compared with controls, EPT infants displayed global reductions in cortical and subcortical gray matter, brainstem, and an increased cerebrospinal fluid volume. Regionally, they showed decreased volumes of all brain tissues, in particular cortical gray matter. Increased volumes of cortical gray and white matter were observed in regions involved in visual processing. Increasing prematurity, intraventricular hemorrhage grade I–II, and patent ductus arteriosus ligation were associated with decreased volumes and had a particular effect on the cerebellum. Concluding, EPT infants without focal brain lesions had an altered brain growth at TEA that particularly affected the gray matter, and varied when it came to the presence of perinatal risk factors. Brain growth gains in EPT infants may be related to a longer extrauterine experience.

Introduction

The brain of the extremely preterm (EPT) infant grows and develops in a completely different way than it would in the womb. In addition, these babies are exposed to extrauterine stimuli, which may affect the processes of brain maturation, and as a result, how the brain gets wired (Lubsen et al. 2011). Altered connectivity of the developing brain may result in cognitive and neuropsychiatric impairment (Skiold et al. 2012; Marret et al. 2013; Serenius et al. 2013). It is, therefore, very important to develop a greater understanding of the structural mechanisms responsible for neurodevelopmental problems, so that effective early intervention programs can be devised.

Most of the earlier studies have focused on infants born later than 27 weeks of gestation, and information on the structural consequences of EPT birth is scarce. Moreover, most of the studies in the field report altered the brain development of preterm children in childhood, whereas our study focuses on the early brain development of EPT infants up to their term-equivalent age (TEA; term age). Similar studies have previously reported significant reductions in deep gray matter volumes (Inder et al. 2005), gray matter structures, and prominent compromise of the white matter (Parikh et al. 2013) in the presence of several clinical complications. Using diffusion magnetic resonance imaging (MRI), we have previously described altered white matter organization in the centrum semiovale and the corpus callosum of EPT infants with diffuse excessive high signal intensities (DEHSIs; Skiold et al. 2010) and a specific pattern of functional connectivity at term age (Fransson et al. 2007, 2011). Taken together, the body of literature suggests different patterns of early brain development in the anatomy and function of the EPT infant. However, brain injuries such as high-grade intraventricular hemorrhage (IVH) and periventricular white matter lesions are very common in these infants, and the relative contribution of prematurity itself is unclear. We therefore studied EPT infants without evidence of focal brain lesions at TEA, based on neonatal ultrasound or MRI, by employing complementary techniques in order to extend previous findings in this population.

Voxel-based morphometry (VBM; Ashburner and Friston 2000) is an automated procedure for quantifying changes from a voxel-by-voxel analysis of MRI data. It has been used to examine the effects of preterm birth during childhood (Kesler et al. 2008; Padilla et al. 2011), adolescence (Nosarti et al. 2008; Nagy et al. 2009), and early adulthood (Allin et al. 2004). Only one study including preterm infants at term age has used a VBM approach to detect gray matter volume differences of the hippocampus (Lodygensky et al. 2008). We aimed to investigate volumetric differences in brain tissues, on a whole-brain level, by using VBM in EPT infants examined at term age. We hypothesized that EPT infants at term age would show differences in brain morphology on a global and regional level when compared with term infants, even without evidence of focal brain lesions being detected by neonatal ultrasound or conventional MRI. We used automated segmentation and VBM improved by a Diffeomorphic Anatomical Registration through Exponentiated Lie algebra (DARTEL) algorithm (Ashburner 2007), respectively. Our secondary objective was to examine the effect of potentially adverse perinatal factors on brain volumes.

Materials and Methods

Subjects

Eligible subjects for this study were all EPT infants (<27 weeks of gestation) born in the Stockholm region between 1 January 2004 and 31 December 2008, who successfully underwent T1- and T2-weighted (w) MRI at term age. We excluded infants with any grade of periventricular leukomalacia (PVL) or IVH grade III and IV on neonatal ultrasound and focal brain lesions (e.g. cysts and infarctions), moderate and severe white matter abnormalities (defined qualitatively) (Inder et al. 2003), or persistent ventricular dilatation on MRI examination at term age. A group of healthy, term-born infants was recruited from the maternity ward and was also scanned at term age. At term age, 190 EPT infants successfully underwent imaging. Of these infants, 19 (9.7%) did not have structural MRI acquisition, meaning that 177 MRI cases were assessed for good quality assurance. The most common reasons for poor quality studies were: the presence of motion artifacts in 62 (35.02%) patients, incomplete coverage of the brain in 38 (21.46%) patients, and blurring of the gray and white matter interfaces in 15 (8.47%) patients. Fifteen (8.47%) MRI cases were excluded for clinical reasons. Two of these patients had noncystic PVL, one had cystic PVL, seven had IVH grades III and IV (six had qualitatively defined white matter abnormalities, with one categorized as severe and five as moderate) (Inder et al. 2003), and five had posthemorrhagic ventriculomegaly. Two of the 15 infants who were excluded were classified as small for gestational age. A control group of 21 healthy, term-born infants underwent MRI, and six (28.57%) of these were excluded due to motion artifacts. Our final sample comprised 47 EPT infants of 177 (26.55%), who all had MRI images suitable for segmentation and volumetric analyses, together with 15 term infants of 21 (71.4%). The regional ethical review board in Stockholm granted ethical approval for the study and written consent was obtained from the parents of the infants.

MRI Data Acquisition

All imaging was performed on a Philips Intera 1.5-T MRI system (Philips International, Amsterdam, The Netherlands). The conventional MRI protocol consisted of a sagittal T1-w turbo spin-echo sequence, an axial inversion recovery sequence, and an axial T2-w sequence. Details of the sequence parameters have previously been published (Skiold et al. 2012). The 3-dimensional T1-w gradient-echo images were acquired with an echo time of 4.6 min, a repetition time of 40 min, a flip angle of 30°, a voxel size of 0.7 × 0.7 × 0.1 mm, and an field of view of 180 mm.

Qualitative Assessment of the MRI Studies

Structural scans (T1- and T2-w images) were assessed by a neuroradiologist to ensure that they were free of gross abnormalities. Qualitative white matter abnormalities were defined based on a previously published scoring system (Inder et al. 2003). This system assessed 5 separate items: Abnormal white matter signal, reduced white matter volume, cystic changes, myelination/thinning of the corpus callosum, and ventricular dilatation. White matter abnormalities were further classified by the composite scores of these 5 categories (ranging from 5 to 15) into: No white matter abnormalities (score 5–6), mild white matter abnormalities (score 7–9), moderate white matter abnormalities (score 10–12), or severe white matter abnormalities (score 13–15).

Automatic Brain Segmentation and Voxel-Based Morphometry

The prior manual steps included reorientation of the original T1-w images in the plane of anterior–posterior commissures and removal of nonbrain tissue components using the Brain Extraction Tool (Smith 2002). Structural images were then segmented into tissue classes using unified segmentation (Ashburner and Friston 2005), as implemented in the “new segment” option of the SPM v8 software, Wellcome Department, University College (London, UK), running on MATLAB v7.5 (MathWorks, Natrick, MA, USA). For guiding segmentation, we used tissue probability maps from preterm infants scanned at term age (Kuklisova-Murgasova et al. 2011). A quantitative evaluation of the segmentations was performed, and the Dice coefficient (Dice 1945) was calculated showing an agreement of 0.87 ± 0.02 for cortical gray matter, 0.86 ± 0.02 for white matter, 0.79 ± 0.03 for deep gray matter, 0.84 ± 0.02 for cerebellum, and 0.83 ± 0.01 for brainstem (see Supplementary Material for further details on automatic brain segmentation and validation).

The segmented brain tissues (see Supplementary Fig. 1) were spatially normalized by using DARTEL. Finally, all images were modulated and smoothed with a full width at half-maximum of 3-mm Gaussian kernel. We used these smoothed brain tissue images to conduct the statistical analyses, as outlined below (see Supplementary Material for further details on VBM). Global brain tissue volumes (in cm3) were extracted from the segmented/normalized/modulated images of each subject with the Easy Volume toolbox (Pernet et al. 2009).

Neonatal Complications and Brain Tissue Volumes

To investigate the effect of the neonatal variables on brain tissue volumes within the preterm group, 2 subgroups were formed based on the presence or absence of the following neonatal variables: Qualitatively defined mild white matter abnormalities; DEHSI; IVH grades I–II; patent ductus arteriosus (PDA); PDA with medical treatment (ibuprofen); retinopathy of prematurity grade >2; and bronchopulmonary dysplasia, defined as the need of oxygen at 36 weeks of gestation and clinical diagnosis of sepsis, with or without positive blood culture. Additional analyses were performed based on: the degree of immaturity (≤25+6 and 26+0–26+6 weeks); any use of mechanical ventilation versus nasal continuous positive airway pressure (nCPAP), and PDA ligation versus PDA with medical treatment. We also studied correlations between brain tissue volumes and days on ventilatory support (mechanical ventilation and nCPAP) in the EPT group.

Statistical Analysis

Studied variables were tested for normality and homogeneity before each analysis. Quantitative data were analyzed with Student's t-test, and qualitative data with Pearson's χ2 or Fisher's exact test. The Mann–Whitney U-test was applied for non-normally distributed data. The statistical significance level was set at P < 0.05. All statistical analyses were computed using the SPSS, version 20.0 (SPSS, Inc., Chicago, IL, USA).

Global Brain Volumes

To assess differences in global volumetric measures between the groups, a general lineal model analysis (multivariate) was performed with all brain volumes as dependent variables and group status as the independent factor. The data were tested to ensure that no data assumptions were violated. The covariates used were total intracranial volume (TIV = all brain tissues, including cerebrospinal fluid [CSF]), to control for generalized scaling effects, and postmenstrual age at the time of the scan. Owing to the fact that variations in CSF volumes could affect the TIV in the EPT infants, we also conducted analyses controlling for the total volume of the cerebral parenchyma (CPAR), including all the brain tissues, but excluding CSF, in addition to postmenstrual age at imaging. For the model, we used a Type III sum of squares, which is invariant with respect to unequal samples. The same model was used to investigate the effect of the neonatal variables on brain tissue volumes between the preterm groups. In this case, analyses were adjusted for gestational age at birth and days on mechanical ventilations as appropriate.

Voxel-Based Morphometry Analysis

For VBM analysis, t-test group comparisons were performed to evaluate the brain volume differences between groups. In the EPT group, a multiple regression analysis was performed to evaluate the relationship between brain tissue volumes and days on ventilatory support (mechanical ventilation and nCPAP). We restricted the statistical analyses to areas with a minimum probability value of 0.25 (absolute threshold masking), avoiding possible edge effects around tissue borders. We analyzed the following contrasts: Patients greater than controls and controls greater than patients. Analyses were adjusted for CPAR. TIV, age at MRI, and gender did not account significantly for variations in the model and were not included in subsequent analyses. We used the same previously defined covariates in the VBM analyses between preterm groups. Only clusters >10 voxels are reported. All statistical images were subjected to family-wise error correction for multiple comparisons. For visualization purposes, the significant clusters were displayed at uncorrected P < 0.001 in order to identify the affected structures and networks.

Results

Table 1 summarizes the neonatal characteristics of the preterm and term cohorts. The neonatal morbidity of the preterm group is reported in Table 2.

Table 1

Characteristics of the study groups

 Preterm (n = 47) Term (n = 15) Statistic (P-value) 
Perinatal 
Gestational age at delivery 25.67 (1.09) 38.85 (0.35) U (<0.001) 
 Range 23.14–26.86 38.42–39.85 
Male/female ratio 21/26 8/7 Fisher's test 0.767 
Birth weight (g) 838.96 (152.83) 3712 (302.88) U (<0.001) 
 Range 528–1161 3075–4100 
Corrected age at scan (weeks) 40.85 (1.02) 40.54 (0.52) t = −1.514 (0.13) 
 Range 38.14–43.28 39.85–41.85 
 Preterm (n = 47) Term (n = 15) Statistic (P-value) 
Perinatal 
Gestational age at delivery 25.67 (1.09) 38.85 (0.35) U (<0.001) 
 Range 23.14–26.86 38.42–39.85 
Male/female ratio 21/26 8/7 Fisher's test 0.767 
Birth weight (g) 838.96 (152.83) 3712 (302.88) U (<0.001) 
 Range 528–1161 3075–4100 
Corrected age at scan (weeks) 40.85 (1.02) 40.54 (0.52) t = −1.514 (0.13) 
 Range 38.14–43.28 39.85–41.85 

Note: Results are expressed as mean (SD) or median (range).

U, Mann–Whitney U-test; t, Student's t-test.

Table 2

Characteristics of the preterm infants

Characteristic Cohort (N = 47) 
Antenatal steroids (%) 47 (100) 
Apgar 5 min, mean (SD), range 7.49 (1.89) (3–10) 
CRIB score, mean (SD), range 5.09 (3.37) (1–11) 
Days on CPAP, days, mean (SD), range 36.91 (13.91) (1–64) 
Days on ventilator, mean (SD), range 11.78 (13.3) (0–47) 
Postnatal corticosteroid therapy (%) 5/43 (11.6) 
Bronchopulmonary dysplasiaa (%) 17/42 (40.4) 
Sepsis (clinical diagnosis) 27/43 (62.7) 
PDA (%) 36/47 (76.6) 
Treated PDA: ibuprofen (%) 21/36 (58.3) 
Treated PDA: surgery (%) 15/36 (41.6) 
Intraventricular haemorrhage grade I (%) 12/47 (25.5) 
Intraventricular haemorrhage grade II (%) 7/47 (14.90) 
No white matter abnormalities (score 5–6) (%) 23/47 (49.0) 
Mild white matter abnormalities (score 7–9) (%) 24/47 (46.80) 
Diffuse and excessive high signal intensities (%) 23/47 (48.9) 
Retinopathy of prematurity stage ≥3 (%) 14/47 (29.8) 
Retinopathy of prematurity stage ≤2  (%) 17/47 (36.1) 
Characteristic Cohort (N = 47) 
Antenatal steroids (%) 47 (100) 
Apgar 5 min, mean (SD), range 7.49 (1.89) (3–10) 
CRIB score, mean (SD), range 5.09 (3.37) (1–11) 
Days on CPAP, days, mean (SD), range 36.91 (13.91) (1–64) 
Days on ventilator, mean (SD), range 11.78 (13.3) (0–47) 
Postnatal corticosteroid therapy (%) 5/43 (11.6) 
Bronchopulmonary dysplasiaa (%) 17/42 (40.4) 
Sepsis (clinical diagnosis) 27/43 (62.7) 
PDA (%) 36/47 (76.6) 
Treated PDA: ibuprofen (%) 21/36 (58.3) 
Treated PDA: surgery (%) 15/36 (41.6) 
Intraventricular haemorrhage grade I (%) 12/47 (25.5) 
Intraventricular haemorrhage grade II (%) 7/47 (14.90) 
No white matter abnormalities (score 5–6) (%) 23/47 (49.0) 
Mild white matter abnormalities (score 7–9) (%) 24/47 (46.80) 
Diffuse and excessive high signal intensities (%) 23/47 (48.9) 
Retinopathy of prematurity stage ≥3 (%) 14/47 (29.8) 
Retinopathy of prematurity stage ≤2  (%) 17/47 (36.1) 

CPAP, continuous positive airway pressure; CRIB, clinical risk index for babies; PDA, patent ductus arteriosus.

aDefined as the need of oxygen at 36 weeks of gestation.

Global Volume Data

Global brain volume measurements are summarized in Table 3. Overall, EPT infants had lower brain volumes than term infants, and we specifically found that cortical gray matter, deep gray matter, and brainstem volumes were significantly lower. The CSF volume was significantly higher in the preterm group. It is noteworthy that the differences in deep gray matter and brainstem volumes remained significant after adjustment for CPAR and the exclusion of infants with mild white matter abnormalities.

Table 3

Cerebral tissue volumes (in cm3) for prematures and term-born infants

Brain volumes Preterm (N = 47) Term (N = 15) Statistic (P-value) volumes unadjusted Statistic (P-value) adjusted for age at scan ICV Statistic (P-value) 
Intracranial volume 465.13 (46.84) 468.89 (25.28) t = −0.50 (0.61)   
Cortical gray matter 
 Mean (SD) 171.18 (18.90) 176.54 (11.91) t = −1.05 (0.29) F = 7.29 (0.009) F = 3.78 (0.05) 
 Range 135.98–214.13 155.21–195.01    
White matter 
 Mean (SD) 146.92 (15.48) 150.06 (10.09) t = −0.73 (0.46) F = 2.52 (0.11) F = 3.78 (0.05) 
 Range 118.97–190.69 131.66–166.09    
Cerebrospinal fluid 
 Mean (SD) 86.46 (10.01) 80.15 (10.37) t = 2.10 (0.03) F = 9.08 (0.004) F = 6.77 (0.01) 
 Range 69.96–115.51 63.56–99.12    
Deep gray matter 
 Mean (SD) 29.12 (2.71) 30.42 (1.82) t = −1.87 (0.06) F = 11.98 (0.001) F = 4.25 (0.04) 
 Range 24.32–37.80 27.47–33.41    
Cerebellum 
 Mean (SD) 26.61 (2.97) 27.35 (2.12) t = 0.12 (0.37) F = 1.95 (0.16) F = 0.17 (0.68) 
 Range 20.12–34.70 24.70–31.89    
Brainstem 
 Mean (SD) 4.70 (0.41) 5.25 (0.31) t = −4.32 (<0.001) F = 45.84 (<0.001) F = 36.62 (<0.001) 
 Range 3.86–5.98 4.73–5.88    
Brain volumes Preterm (N = 47) Term (N = 15) Statistic (P-value) volumes unadjusted Statistic (P-value) adjusted for age at scan ICV Statistic (P-value) 
Intracranial volume 465.13 (46.84) 468.89 (25.28) t = −0.50 (0.61)   
Cortical gray matter 
 Mean (SD) 171.18 (18.90) 176.54 (11.91) t = −1.05 (0.29) F = 7.29 (0.009) F = 3.78 (0.05) 
 Range 135.98–214.13 155.21–195.01    
White matter 
 Mean (SD) 146.92 (15.48) 150.06 (10.09) t = −0.73 (0.46) F = 2.52 (0.11) F = 3.78 (0.05) 
 Range 118.97–190.69 131.66–166.09    
Cerebrospinal fluid 
 Mean (SD) 86.46 (10.01) 80.15 (10.37) t = 2.10 (0.03) F = 9.08 (0.004) F = 6.77 (0.01) 
 Range 69.96–115.51 63.56–99.12    
Deep gray matter 
 Mean (SD) 29.12 (2.71) 30.42 (1.82) t = −1.87 (0.06) F = 11.98 (0.001) F = 4.25 (0.04) 
 Range 24.32–37.80 27.47–33.41    
Cerebellum 
 Mean (SD) 26.61 (2.97) 27.35 (2.12) t = 0.12 (0.37) F = 1.95 (0.16) F = 0.17 (0.68) 
 Range 20.12–34.70 24.70–31.89    
Brainstem 
 Mean (SD) 4.70 (0.41) 5.25 (0.31) t = −4.32 (<0.001) F = 45.84 (<0.001) F = 36.62 (<0.001) 
 Range 3.86–5.98 4.73–5.88    

Note: Unadjusted brain volumes were compared by using a Student's t-test. Unadjusted volumes were compared by using a multivariate analysis of covariance.

ICV, intracranial volume (sum of all brain tissues); CPAR, cerebral parenchyma (excluding cerebrospinal fluid).

Whole-Brain Voxel-Based Morphometry Analysis

The results of this section are presented in Supplementary Table 1. Compared with the term group, EPT infants exhibited significant regional volume changes in all brain tissues. The cortical gray matter reductions were bilaterally distributed in the preterm group (Fig. 1A). These reductions predominantly occurred in the cortical temporal lobe, extending superiorly to the perirolandic cortex, inferiorly to the orbito-frontal regions, and medially to the entorhinal cortex. Regions in the left insula also showed a decreased volume. The results were not altered when infants treated with postnatal steroids (n = 5) were excluded from the analysis. Smaller clusters of white matter decrements were observed in areas adjacent to the gray matter decreases, primarily affecting the temporal cortex and, to a minor extent, the angular gyrus and perirolandic area bilaterally (Fig. 1B). Notably, larger gray matter and white matter volumes were detected bilaterally in the occipital, parietal, and frontal cortices (Fig. 1C,D). The preterm group also displayed areas of decreased volume in the deep gray matter, cerebellum, and brainstem (Fig. 2).

Figure 1.

Regional volumetric differences (warm colors) in the EPT infants compared with term infants (A) gray matter decreases, (B) white matter decreases, (C) gray matter increases, and (D) white matter increases. The color bar represents the t-score. Display orientation: right = right. Differences are mapped on the render of one preterm infant.

Figure 1.

Regional volumetric differences (warm colors) in the EPT infants compared with term infants (A) gray matter decreases, (B) white matter decreases, (C) gray matter increases, and (D) white matter increases. The color bar represents the t-score. Display orientation: right = right. Differences are mapped on the render of one preterm infant.

Figure 2.

Sagittal, coronal, and axial slices showing regional volumetric differences in the EPT infants, compared with term infants, in (A) deep gray matter, (B) cerebellum, and (C) brainstem. The color bar represents the t-score. Display orientation: right = right. Results are superimposed on the smoothed corresponding image of one preterm infant.

Figure 2.

Sagittal, coronal, and axial slices showing regional volumetric differences in the EPT infants, compared with term infants, in (A) deep gray matter, (B) cerebellum, and (C) brainstem. The color bar represents the t-score. Display orientation: right = right. Results are superimposed on the smoothed corresponding image of one preterm infant.

Perinatal Risk Factors: Global and Regional Brain Volumes

The most immature infants, those with low-grade IVH and also with PDA who needed surgical ligation, had significant reduced global and regional brain volumes (Supplementary Tables 2 and 3) in several regions (Fig. 3). Notably, cerebellar growth was affected in the presence of all of these perinatal risk factors. The neonatal variables that significantly differed between preterm groups were used as covariates in all analyses (Supplementary Table 4). Of note, areas of increased cortical gray matter were detected in the medically PDA treated group, when compared with the group without PDA (P < 0.005). Although we could not find global brain volume differences between infants on mechanical ventilation in comparison with those using nCPAP, a cluster of decreased cortical gray matter volume in the left primary motor cortex (cluster = 475 voxels, P = 0.04) was identified on a regional level (not shown). Significant negative correlations were observed between days on mechanical ventilation and brain volumes in deep gray matter (r = 0.51, P < 0.001), cerebellum (r = 0.57, P < 0.001), and brainstem (r = 0.53, P < 0.001) (Supplementary Fig. 2). With respect to other medical conditions, we could not find any differences in global or regional level brain volumes between the preterm groups.

Figure 3.

Sagittal, coronal, and axial slices showing regional volumetric differences between preterm infants under specific neonatal conditions. The color bar represents the t-score. Display orientation: right = right. Results are superimposed on the smoothed corresponding image of one preterm infant.

Figure 3.

Sagittal, coronal, and axial slices showing regional volumetric differences between preterm infants under specific neonatal conditions. The color bar represents the t-score. Display orientation: right = right. Results are superimposed on the smoothed corresponding image of one preterm infant.

Discussion

The main findings of this study are that when the EPT infants were compared with the term infants, they showed significant global reductions in brain tissue, accompanied by a significant increase in CSF volume. On a regional level, the preterm infants showed decreased volumes of all brain tissues, in particular cortical gray matter. An unexpected finding of the study is that the EPT infants displayed increased cortical gray matter and white matter involving vision-related brain areas. Finally, increasing prematurity, low-grade IVH, and PDA ligation were associated with different patterns of brain volume reductions and had a particular effect on the cerebellum. Our study extends previous findings and supports the notion that EPT birth induces deviations in the developmental trajectory of the brain at term age, even in the absence of focal brain pathology.

Global Brain Volumes

We found decreased global volumes of cortical gray matter, deep gray matter, and brainstem. This is similar to a previous report (Parikh et al. 2013), which also found smaller global brain volumes, but with a major involvement of the white matter. However, this might be due to the fact that their study population had more clinical complications than ours. Moreover, 25% of the infants in that study were small for their gestational age, a condition that is known to affect white matter maturation in preterm infants at term age (Lepomaki et al. 2013). In the current study, major volume reductions were found in the deep gray matter and brainstem of the EPT infants. The volumetric differences in the deep gray matter persisted even when conventional MRI sequences failed to identify any white matter abnormality in infants. Our results confirm previous findings that the deep gray matter is specifically vulnerable after preterm birth and is particularly affected in more immature infants (Inder et al. 2005). In this study, EPT infants had significant reductions in the whole-brainstem volume. At this early age, the brainstem is part of a vertical-integrative hierarchical system, along with the limbic and cortical systems, and is involved in the process of regulatory functions, such as arousal, attention, and emotion (Geva and Feldman 2008).

The global volume reductions found in our study were probably not due to acute or hypoxic injury, as there were no signs of brain abnormalities caused by these conditions. It is more than likely that the present results can be explained by the extreme preterm infants (mean gestational age of 25.6 weeks) being born during a critical period of brain development, when synaptic density neuronal wiring is increasing in all the cortical regions (Zecevic 1998), and the thalamo-cortical projections are waiting within the subplate zone to reach the cortical plate (Kostovic and Judas 2010). In addition, at the time of EPT birth, the many cortical interneurons in the human fetus are been generated in the proliferative zones and are still migrating to the cerebral cortex (Letinic and Rakic 2001). Altogether, the EPT birth per se probably leads to altered connections at multiple anatomical levels, from the cortex to the deep gray matter and brainstem, with subsequent developmental disturbances and trophic consequences reflected by volume alterations (Volpe 2009a, 2009b; Kostovic and Judas 2010; Ball et al. 2012).

Regional Analyses: Voxel-Based Morphometry

We found that when we compared the EPT infants with the term infants in the control group, they showed extensive regions of reduced gray matter in several brain areas that have previously been described in more mature infants studied in childhood and adolescence. These areas included: the cortical temporal lobe bilaterally (Kesler et al. 2008; Nosarti et al. 2008; Nagy et al. 2009), precentral and postcentral gyri (Peterson et al. 2003; Allin et al. 2004; Kesler et al. 2008); orbito-frontal cortex (Thompson et al. 2007; Nagy et al. 2009; Ball et al. 2012); amygdala (Peterson et al. 2000; Kesler et al. 2008); parahippocampal gyrus (Kesler et al. 2008); hippocampus (Peterson et al. 2000; Kesler et al. 2008; Nagy et al. 2009), and left insula (Nosarti et al. 2008). The additional involvement of subcortical gray matter sites, including the deep gray matter (an interface between neocortical and lower-level systems), cerebellum, and brainstem, confirms the increase susceptibility of the gray matter in the preterm brain (Dean et al. 2013; Malik et al. 2013).

Structures involved in auditory, language, and somatosensory processing were recognized. We also identified a structure that has not previously been reported in EPT infants at term age, namely the entorhinal cortex, a gateway between neocortical association areas and the hippocampal system (Fransen 2005). In adolescents born preterm, entorhinal thinning has been associated with impaired cognitive function as a result of aberrant cortical maturation (Skranes et al. 2012). Further studies should focus on the entorhinal cortex and its correlation with cognition in EPT samples. The decreased regional volumes found in our study may be due to compromised neurogenesis (Malik et al. 2013), disturbances of the dendritic arbor, and synapse formation of cortical neurons (Dean et al. 2013).

Increased volumes were detected in regions known to be involved in visual processing, contrary to previous studies (Peterson et al. 2003; Shah et al. 2006; Thompson et al. 2007). However, the infants in these studies were more mature than the preterm infants included in our study and experienced several clinical complications. The nature of the increase in gray and white matter is difficult to interpret as the VBM approach provides a measure of volume, but provides no information on the cytoarchitectural structure. Increases in gray and white matter volumes may reflect an abnormal or delayed pruning program, as has been previously suggested (Nosarti et al. 2008). However, the visual cortex represents a particular region of the brain that matures earlier than other areas of the cortex (Tzarouchi et al. 2009) and has a functional relevance in EPT infants at term age (Fransson et al. 2007,, 2011). In this respect, increased volumes in our preterm cohort may correspond to advanced maturational aspects in the visual regions, due to longer extrauterine visual experiences during a critical period, compared with term infants. In fact, both classic (Wiesel 1982; Bourgeois et al. 1989) and modern studies (Tsuneishi and Casaer 2000; Gimenez et al. 2008) support the theory that neural activity driven by visual experience is essential for shaping the early rudimentary cortical connectivity patterns (Penn and Shatz 1999). Further studies that follow up functional aspects are needed to define the developmental correlates of these findings.

Perinatal Risk Factors

In addition to the degree of immaturity, the presence of IVH grade I–II and surgically ligated PDA were associated with reduced brain growth. The regional gray matter was also affected by the severity of respiratory illness. Overall, brain volume reductions preferentially affected gray matter structures and, in particular, the cerebellum, as mentioned above. The other perinatal factors that were studied had no significant effect on cerebral structure at term age.

In this study, we included infants with IVH grade I–II as these hemorrhages are usually considered to be benign. Our findings of reduced brain volumes are in agreement with previous MRI studies (Vasileiadis et al. 2004) and may reflect the influence of the low-grade IVH on developmental processes related to suppression of cell proliferation (Del Bigio 2011) and microglia activation (Supramaniam et al. 2013). These findings may help to explain the impaired developmental outcomes reported in preterm infants with low-grade IVH at different ages (Klebermass-Schrehof et al. 2012).

The adverse effect of PDA ligation may be due to disturbed cerebral circulation. Infants who needed surgical ligation faced a higher risk of affected brain growth than those who only received medical treatment. There seem to be adverse effects of the PDA ligation per se, where intraoperative or postoperative physiological alterations may affect cerebral hemodynamics (El-Khuffash et al. 2013), potentially causing further injury to an already vulnerable brain. It is, however, surprising to note that the infants in our cohort who only received medical treatment (ibuprofen) for PDA had preserved brain volumes compared with those without. We speculate that the anti-inflammatory effect of ibuprofen might be protective for brain growth, which is in line with previous findings where ibuprofen showed neuroprotective effects attenuating neuronal damage (Wixey et al. 2012).

The neuropathological basis of the decreased regional cerebellar volume in our study still needs to be elucidated. Of note, the gestational age of the EPT neonates included in our research corresponds to a phase of rapid cerebellar development, representing a period of particular vulnerability. It is, therefore, likely that the suppression of neuronal proliferation (Volpe 2009a, 2009b), trophic interactions (Tam et al. 2011), and a compromised cerebellar perfusion (Limperopoulos et al. 2005) might specifically affect the growth of the cerebellum. Impaired cerebellar growth accounts for subsequent developmental problems, which indicate the need for heightened clinical surveillance and early target interventions. Remarkably, the presence of DEHSI detected in 48.9% of our preterm cohort at term age was not associated with brain volume alterations. We have previously found that the presence of DEHSI in EPT infants at term age was not related to abnormal neurologic outcome at 30 months (Skiold et al. 2012). These results suggested that DEHSI is a prematurity-related transient phenomenon.

Our study benefits from including a unique cohort of EPT infants without major brain alterations on neonatal ultrasound or MRI at term age and a mean gestational age lower than previously published cohort studies. However, some methodological concerns must be considered. We are aware that tissue segmentation in preterm infants at term is a challenging task, owing to the developmental characteristics of the preterm brain. Furthermore, the segmentation is limited in smaller cerebral structures, owing to the smaller volumes and a lower signal-to-noise ratio than in older infants. To minimize these limitations, we only used good quality T1-w images. For guiding segmentation, we used a larger number of tissue probability maps, together with an extraclass tissue map for background, to provide better modeling of CSF and other nonbrain voxels and further to improve the tissue classification. Additionally, the tissue probability maps used here were based on a sample of preterm neonates scanned at term age. We also used the DARTEL algorithm to provide a better intersubject registration.

In conclusion, we have shown that at TEA, and in the absence of focal brain lesions, EPT infants have an altered global and regional brain growth pattern, involving decreases and increases of brain tissue volumes. Gray and white matter gains in regions known to be involved in visual processing suggest maturational brain changes driven by precocious visual experiences in EPT infants. The degree of immaturity, IVH grade I–II, and surgically ligated PDA were associated with global and regional decreased brain growth that preferentially affected gray matter structures, in particular the cerebellum. Our study suggests that structural alterations may be used as structural markers to identify children at risk earlier, and that this may prevent adverse outcomes following EPT birth. Longitudinal studies, including functional follow-up data, would help to determine the significance of the volumetric brain alterations found in the EPT infants in our study.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/.

Funding

This work was supported by the following grants: Swedish Medical Research Council (2011–3981), the regional agreement on medical training and clinical research (ALF 20120450) between Stockholm County Council and Karolinska Institutet, Marianne and Marcus Wallenberg Foundation (MMW 2011.0085), EU Seventh Framework (223767), Linnea and Josef Carlssońs Foundation, Swedish Order of Freemasons in Stockholm, Swedish Medical Society, Swedish Brain Foundation, and Sällskapet Barnavård. N.P. was supported by a Sara Borrell post-doctoral fellowship (CD09/00263), Instituto de Salud Carlos III, Spain.

Notes

We thank the families and infants who took part in the study, together with the MR physics group and MR radiology staff at Astrid Lindgren Children's Hospital for their valuable technical assistance. Conflict of Interest: None declared.

References

Allin
M
Henderson
M
Suckling
J
Nosarti
C
Rushe
T
Fearon
P
Stewart
AL
Bullmore
ET
Rifkin
L
Murray
R
.
2004
.
Effects of very low birthweight on brain structure in adulthood
.
Dev Med Child Neurol
 .
46
:
46
53
.
Ashburner
J
.
2007
.
A fast diffeomorphic image registration algorithm
.
Neuroimage
 .
38
:
95
113
.
Ashburner
J
Friston
KJ
.
2005
.
Unified segmentation
.
Neuroimage
 .
26
:
839
851
.
Ashburner
J
Friston
KJ
.
2000
.
Voxel-based morphometry—the methods
.
Neuroimage
 .
11
:
805
821
.
Ball
G
Boardman
JP
Rueckert
D
Aljabar
P
Arichi
T
Merchant
N
Gousias
IS
Edwards
AD
Counsell
SJ
.
2012
.
The effect of preterm birth on thalamic and cortical development
.
Cereb Cortex
 .
22
:
1016
1024
.
Bourgeois
JP
Jastreboff
PJ
Rakic
P
.
1989
.
Synaptogenesis in visual cortex of normal and preterm monkeys: evidence for intrinsic regulation of synaptic overproduction
.
Proc Natl Acad Sci USA
 .
86
:
4297
4301
.
Dean
JM
McClendon
E
Hansen
K
Azimi-Zonooz
A
Chen
K
Riddle
A
Gong
X
Sharifnia
E
Hagen
M
Ahmad
T
et al
2013
.
Prenatal cerebral ischemia disrupts MRI-defined cortical microstructure through disturbances in neuronal arborization
.
Sci Transl Med
 .
5
:
168ra167
.
Del Bigio
MR
.
2011
.
Cell proliferation in human ganglionic eminence and suppression after prematurity-associated haemorrhage
.
Brain
 .
134
:
1344
1361
.
Dice
LR
.
1945
.
Measures of the amount of ecologic association between species
.
Ecology
 .
26
:
297
302
.
El-Khuffash
AF
Jain
A
McNamara
PJ
.
2013
.
Ligation of the patent ductus arteriosus in preterm infants: understanding the physiology
.
J Pediatr
 .
162
:
1100
1106
.
Fransen
E
.
2005
.
Functional role of entorhinal cortex in working memory processing
.
Neural Netw
 .
18
:
1141
1149
.
Fransson
P
Aden
U
Blennow
M
Lagercrantz
H
.
2011
.
The functional architecture of the infant brain as revealed by resting-state fMRI
.
Cereb Cortex
 .
21
:
145
154
.
Fransson
P
Skiold
B
Horsch
S
Nordell
A
Blennow
M
Lagercrantz
H
Aden
U
.
2007
.
Resting-state networks in the infant brain
.
Proc Natl Acad Sci USA
 .
104
:
15531
15536
.
Geva
R
Feldman
R
.
2008
.
A neurobiological model for the effects of early brainstem functioning on the development of behavior and emotion regulation in infants: implications for prenatal and perinatal risk
.
J Child Psychol Psychiatry
 .
49
:
1031
1041
.
Gimenez
M
Miranda
MJ
Born
AP
Nagy
Z
Rostrup
E
Jernigan
TL
.
2008
.
Accelerated cerebral white matter development in preterm infants: a voxel-based morphometry study with diffusion tensor MR imaging
.
Neuroimage
 .
41
:
728
734
.
Inder
TE
Warfield
SK
Wang
H
Huppi
PS
Volpe
JJ
.
2005
.
Abnormal cerebral structure is present at term in premature infants
.
Pediatrics
 .
115
:
286
294
.
Inder
TE
Wells
SJ
Mogridge
NB
Spencer
C
Volpe
JJ
.
2003
.
Defining the nature of the cerebral abnormalities in the premature infant: a qualitative magnetic resonance imaging study
.
J Pediatr
 .
143
:
171
179
.
Kesler
SR
Reiss
AL
Vohr
B
Watson
C
Schneider
KC
Katz
KH
Maller-Kesselman
J
Silbereis
J
Constable
RT
Makuch
RW
et al
2008
.
Brain volume reductions within multiple cognitive systems in male preterm children at age twelve
.
J Pediatr
 .
152
:
513
520
,
520 e511
.
Klebermass-Schrehof
K
Czaba
C
Olischar
M
Fuiko
R
Waldhoer
T
Rona
Z
Pollak
A
Weninger
M
.
2012
.
Impact of low-grade intraventricular hemorrhage on long-term neurodevelopmental outcome in preterm infants
.
Childs Nerv Syst
 .
28
:
2085
2092
.
Kostovic
I
Judas
M
.
2010
.
The development of the subplate and thalamocortical connections in the human foetal brain
.
Acta Paediatr
 .
99
:
1119
1127
.
Kuklisova-Murgasova
M
Aljabar
P
Srinivasan
L
Counsell
SJ
Doria
V
Serag
A
Gousias
IS
Boardman
JP
Rutherford
MA
Edwards
AD
et al
2011
.
A dynamic 4D probabilistic atlas of the developing brain
.
Neuroimage
 .
54
:
2750
2763
.
Lepomaki
V
Matomaki
J
Lapinleimu
H
Lehtonen
L
Haataja
L
Komu
M
Parkkola
R
.
2013
.
Effect of antenatal growth on brain white matter maturation in preterm infants at term using tract-based spatial statistics
.
Pediatr Radiol
 .
43
:
80
85
.
Letinic
K
Rakic
P
.
2001
.
Telencephalic origin of human thalamic GABAergic neurons
.
Nat Neurosci
 .
4
:
931
936
.
Limperopoulos
C
Soul
JS
Gauvreau
K
Huppi
PS
Warfield
SK
Bassan
H
Robertson
RL
Volpe
JJ
du Plessis
AJ
.
2005
.
Late gestation cerebellar growth is rapid and impeded by premature birth
.
Pediatrics
 .
115
:
688
695
.
Lodygensky
GA
Seghier
ML
Warfield
SK
Tolsa
CB
Sizonenko
S
Lazeyras
F
Huppi
PS
.
2008
.
Intrauterine growth restriction affects the preterm infant's hippocampus
.
Pediatr Res
 .
63
:
438
443
.
Lubsen
J
Vohr
B
Myers
E
Hampson
M
Lacadie
C
Schneider
KC
Katz
KH
Constable
RT
Ment
LR
.
2011
.
Microstructural and functional connectivity in the developing preterm brain
.
Semin Perinatol
 .
35
:
34
43
.
Malik
S
Vinukonda
G
Vose
LR
Diamond
D
Bhimavarapu
BB
Hu
F
Zia
MT
Hevner
R
Zecevic
N
Ballabh
P
.
2013
.
Neurogenesis continues in the third trimester of pregnancy and is suppressed by premature birth
.
J Neurosci
 .
33
:
411
423
.
Marret
S
Marchand-Martin
L
Picaud
JC
Hascoet
JM
Arnaud
C
Roze
JC
Truffert
P
Larroque
B
Kaminski
M
Ancel
PY
.
2013
.
Brain injury in very preterm children and neurosensory and cognitive disabilities during childhood: the EPIPAGE cohort study
.
PLoS ONE
 .
8
:
e62683
.
Nagy
Z
Ashburner
J
Andersson
J
Jbabdi
S
Draganski
B
Skare
S
Bohm
B
Smedler
AC
Forssberg
H
Lagercrantz
H
.
2009
.
Structural correlates of preterm birth in the adolescent brain
.
Pediatrics
 .
124
:
e964
e972
.
Nosarti
C
Giouroukou
E
Healy
E
Rifkin
L
Walshe
M
Reichenberg
A
Chitnis
X
Williams
SC
Murray
RM
.
2008
.
Grey and white matter distribution in very preterm adolescents mediates neurodevelopmental outcome
.
Brain
 .
131
:
205
217
.
Padilla
N
Falcon
C
Sanz-Cortes
M
Figueras
F
Bargallo
N
Crispi
F
Eixarch
E
Arranz
A
Botet
F
Gratacos
E
.
2011
.
Differential effects of intrauterine growth restriction on brain structure and development in preterm infants: a magnetic resonance imaging study
.
Brain Res
 .
1382
:
98
108
.
Parikh
NA
Lasky
RE
Kennedy
KA
McDavid
G
Tyson
JE
.
2013
.
Perinatal factors and regional brain volume abnormalities at term in a cohort of extremely low birth weight infants
.
PLoS ONE
 .
8
:
e62804
.
Penn
AA
Shatz
CJ
.
1999
.
Brain waves and brain wiring: the role of endogenous and sensory-driven neural activity in development
.
Pediatr Res
 .
45
:
447
458
.
Pernet
C
Andersson
J
Paulesu
E
Demonet
JF
.
2009
.
When all hypotheses are right: a multifocal account of dyslexia
.
Hum Brain Mapp
 .
30
:
2278
2292
.
Peterson
BS
Anderson
AW
Ehrenkranz
R
Staib
LH
Tageldin
M
Colson
E
Gore
JC
Duncan
CC
Makuch
R
Ment
LR
.
2003
.
Regional brain volumes and their later neurodevelopmental correlates in term and preterm infants
.
Pediatrics
 .
111
:
939
948
.
Peterson
BS
Vohr
B
Staib
LH
Cannistraci
CJ
Dolberg
A
Schneider
KC
Katz
KH
Westerveld
M
Sparrow
S
Anderson
AW
et al
2000
.
Regional brain volume abnormalities and long-term cognitive outcome in preterm infants
.
JAMA
 .
284
:
1939
1947
.
Serenius
F
Kallen
K
Blennow
M
Ewald
U
Fellman
V
Holmstrom
G
Lindberg
E
Lundqvist
P
Marsal
K
Norman
M
et al
2013
.
Neurodevelopmental outcome in extremely preterm infants at 2.5 years after active perinatal care in Sweden
.
JAMA
 .
309
:
1810
1820
.
Shah
DK
Guinane
C
August
P
Austin
NC
Woodward
LJ
Thompson
DK
Warfield
SK
Clemett
R
Inder
TE
.
2006
.
Reduced occipital regional volumes at term predict impaired visual function in early childhood in very low birth weight infants
.
Invest Ophthalmol Vis Sci
 .
47
:
3366
3373
.
Skiold
B
Horsch
S
Hallberg
B
Engstrom
M
Nagy
Z
Mosskin
M
Blennow
M
Aden
U
.
2010
.
White matter changes in extremely preterm infants, a population-based diffusion tensor imaging study
.
Acta Paediatr
 .
99
:
842
849
.
Skiold
B
Vollmer
B
Bohm
B
Hallberg
B
Horsch
S
Mosskin
M
Lagercrantz
H
Aden
U
Blennow
M
.
2012
.
Neonatal magnetic resonance imaging and outcome at age 30 months in extremely preterm infants
.
J Pediatr
 .
160
:
559
566
;
e551
.
Skranes
J
Lohaugen
GC
Evensen
KA
Indredavik
MS
Haraldseth
O
Dale
AM
Brubakk
AM
Martinussen
M
.
2012
.
Entorhinal cortical thinning affects perceptual and cognitive functions in adolescents born preterm with very low birth weight (VLBW)
.
Early Hum Dev
 .
88
:
103
109
.
Smith
SM
.
2002
.
Fast robust automated brain extraction
.
Hum Brain Mapp
 .
17
:
143
155
.
Supramaniam
V
Vontell
R
Srinivasan
L
Wyatt-Ashmead
J
Hagberg
H
Rutherford
M
.
2013
.
Microglia activation in the extremely preterm human brain
.
Pediatr Res
 .
73
:
301
309
.
Tam
EW
Miller
SP
Studholme
C
Chau
V
Glidden
D
Poskitt
KJ
Ferriero
DM
Barkovich
AJ
.
2011
.
Differential effects of intraventricular hemorrhage and white matter injury on preterm cerebellar growth
.
J Pediatr
 .
158
:
366
371
.
Thompson
DK
Warfield
SK
Carlin
JB
Pavlovic
M
Wang
HX
Bear
M
Kean
MJ
Doyle
LW
Egan
GF
Inder
TE
.
2007
.
Perinatal risk factors altering regional brain structure in the preterm infant
.
Brain
 .
130
:
667
677
.
Tsuneishi
S
Casaer
P
.
2000
.
Effects of preterm extrauterine visual experience on the development of the human visual system: a flash VEP study
.
Dev Med Child Neurol
 .
42
:
663
668
.
Tzarouchi
LC
Astrakas
LG
Xydis
V
Zikou
A
Kosta
P
Drougia
A
Andronikou
S
Argyropoulou
MI
.
2009
.
Age-related grey matter changes in preterm infants: an MRI study
.
Neuroimage
 .
47
:
1148
1153
.
Vasileiadis
GT
Gelman
N
Han
VK
Williams
LA
Mann
R
Bureau
Y
Thompson
RT
.
2004
.
Uncomplicated intraventricular hemorrhage is followed by reduced cortical volume at near-term age
.
Pediatrics
 .
114
:
e367
e372
.
Volpe
JJ
.
2009a
.
Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances
.
Lancet Neurol
 .
8
:
110
124
.
Volpe
JJ
.
2009b
.
Cerebellum of the premature infant: rapidly developing, vulnerable, clinically important
.
J Child Neurol
 .
24
:
1085
1104
.
Wiesel
TN
.
1982
.
Postnatal development of the visual cortex and the influence of environment
.
Nature
 .
299
:
583
591
.
Wixey
JA
Reinebrant
HE
Buller
KM
.
2012
.
Post-insult ibuprofen treatment attenuates damage to the serotonergic system after hypoxia-ischemia in the immature rat brain
.
J Neuropathol Exp Neurol
 .
71
:
1137
1148
.
Zecevic
N
.
1998
.
Synaptogenesis in layer I of the human cerebral cortex in the first half of gestation
.
Cereb Cortex
 .
8
:
245
252
.