TY - JOUR
AU - Knirsch,, Walter
AB - Abstract Congenital heart defects are the most common congenital anomalies, accounting for a third of all congenital anomaly cases. While surgical correction dramatically improved survival rates, the lag behind normal neurodevelopment appears to persist. Deficits in higher cognitive functions are particularly common, including developmental delay in communication and oral-motor apraxia. It remains unclear whether the varying degree of cognitive developmental delay is reflected in variability in brain growth patterns. To answer this question, we aimed to investigate whether the rate of regional brain growth is correlated with later life neurodevelopment. Forty-four newborns were included in our study, of whom 33 were diagnosed with dextro-transposition of the great arteries and 11 with other forms of severe congenital heart defects. During the first month of life, neonates underwent corrective or palliative cardiovascular bypass surgery, pre- and postoperative cerebral MRI were performed 18.7 ± 7.03 days apart. MRI was performed in natural sleep on a 3.0 T scanner using an 8-channel head coil, fast spin-echo T2-weighted anatomical sequences were acquired in three planes. Based on the principles of deformation-based morphometry, we calculated brain growth rate maps reflecting average daily growth occurring between pre- and postoperative brain images. An explorative, whole-brain, threshold-free cluster enhancement analysis revealed strong correlation between the growth rate of the Heschl’s gyrus, anterior planum temporale and language score at 12 months of age, corrected for demographic variables (P = 0.018, t = 5.656). No significant correlation was found between brain growth rates and motor or cognitive scores. Post hoc analysis showed that the length of hospitalization interacted with this correlation, longer hospitalization resulted in faster enlargement of the internal CSF spaces. Our longitudinal cohort study provides evidence for the early importance of left-dominant perisylvian regions in auditory and language development before direct postnatal exposure to native language. In congenital heart disease patients, the perioperative period results in a critical variability of brain growth rate in this region, which is a reliable neural correlate of language development at 1 year of age. language development, neurodevelopmental impairment, brain growth, magnetic resonance imaging Introduction Late gestational and early postnatal neurodevelopment comprises important developmental processes, such as the rapid increase of cerebral volume and maturation of white matter. There is compelling evidence that congenital heart defects (CHD) adversely affect neurodevelopment during this period. The aetiology of such cerebral injury is most likely multifactorial and not fully understood (McQuillen et al., 2010). As a consequence of altered cerebral development, cognitive development may be impaired in children with more severe forms of CHD (Snookes et al., 2010; Owen et al., 2014; von Rhein et al., 2015; Naef et al., 2017). Furthermore, CHD can lead to a variety of neurodevelopmental impairments: in infancy, they are characterized by abnormalities of muscular tone, feeding difficulties and developmental delays in major motor milestones (Limperopoulos et al., 2002). A variety of neurodevelopmental deficits manifest at later ages and may persist into adolescence and adulthood (Schaefer et al., 2013). Higher cognitive functions are particularly affected during later development (Bellinger et al., 2009; Cassidy et al., 2015), and developmental delay in communication and oral-motor apraxia also occur with a high prevalence (Bellinger et al., 1999; Hovels-Gurich et al., 2002; Kirshbom et al., 2002). Children with CHD often score in the low-normal range on language tests, with 25% classified as at risk for both expressive and receptive language difficulties (Hovels-Gurich et al., 2006). Chronic hypoxemia due to cyanotic cardiac defects is more strongly associated with dysfunction of speech and language (Hovels-Gurich et al., 2008). The early identification of infants at risk for impaired neurocognitive development represents an important priority in clinical counselling of CHD patients and may lead to early therapeutic interventions. Magnetic resonance imaging (MRI) studies have the potential to unravel early structural correlates of adverse neurocognitive outcome in CHD, such as smaller cerebral volumes in utero and in later life. Previous works found a correlation between cerebral volume prior to neonatal surgery and the degree of neurobehavioral development (Andropoulos et al., 2014; Owen et al., 2014). The global reduction of cerebral volume has been correlated with functional outcome (von Rhein et al., 2014). The majority of studies could not unambiguously pinpoint a predilection of pathological changes, such as white matter injury (Miller et al., 2007; Licht et al., 2009; Ortinau et al., 2012b; Volpe, 2014), or delayed cortical gray matter development (Licht et al., 2009; Ortinau et al., 2012b; Claessens et al., 2016) to any of the cerebral lobes or smaller subdivisions. The additive effect of white matter injury and delayed cerebral maturation may explain why CHD infants lag behind normal neurodevelopment, and the varying degree of such injuries could presumably lead to variability in brain growth patterns. Understanding whether this variability in structural maturation is related to the neurocognitive outcomes requires further evidence. Our purpose was to reveal possible morphological correlates of neurocognitive development based on longitudinally acquired MRI in newborns who underwent corrective surgery for complex CHD. Most studies that used neuroimaging methods in CHD focused on a single time point during pregnancy or child development (Limperopoulos et al., 2010; von Rhein et al., 2014, 2015), while others used repeated ultrasonography (Turan et al., 2017), MRI (Ortinau et al., 2012a, b) or head circumference measurement to characterize impaired developmental trajectories (Licht et al., 2009). We assume that a single time-point approach may not be sensitive enough to characterize a critical period during which cerebral developmental is presumed to arrest or slowed. In contrast, repeated, longitudinal studies have the potential to link individual variability in postoperative cerebral growth rates to later cognitive development. Materials and methods Patient population Patients for the longitudinal MRI analysis were sampled from an ongoing prospective cohort study that investigates neurodevelopmental outcome in infants who were operated for CHD within the first month after birth. Between December 2009 and August 2016, 78 infants with a severe type of CHD met the original study criteria. Inclusion criteria for the current work were the availability of pre- and postoperative MRI with sufficient image quality and availability of neurodevelopmental assessment at 1 year of age. Neonates with a suspected or confirmed genetic disorder or syndrome were excluded. Forty-four infants met these criteria and were included. Newborns were diagnosed with dextro-transposition of the great arteries (n = 33) planned for biventricular repair, and non-transposition of the great arteries pathologies consisting of a single ventricle undergoing staged palliation until Fontan procedure (n = 5), interrupted aortic arch or coarctation of the aorta or transverse aortic arch hypoplasia (n = 4). One case of complete unbalanced atrio-ventricular septum defect, one patient with perimembranous ventricular septum defect with open foramen ovale and one infant with pulmonary atresia with ventricular septum defect were included. Subjects were enrolled in the study after birth. Subject demographics are detailed in Table 1. Statistics on the presence of stroke, white matter injury or T2-hyperintensities in the pre- or postoperative images are summarized in Supplementary Table 1. Table 1 Demographic characteristics of the study population Total population Gender, male/female 34/10 Gestational weight at birth, g 3364 ± 446 (2560–4270) Head circumference at birth, cm 34.49 ± 1.2 (32–37) Gestational age at birth, weeks 39.66 ± 1.2 (37–41.7) Gestational age at first MRI, weeks 40.78 ± 2.18 (37.6–51.6) Time between first and second MRI scan, days 18.48 ± 6.81 (7–38) Socioeconomic status 9.09 ± 2.1 (5–12) Days of hospitalization 32.68 ± 11.43 (13–73) Bayley CCS 107.73 ± 14.32 (60–130) Bayley LCS 93.49 ± 14.076 (65–132) Bayley MCS 93.77 ± 13.51 (46–130) Total population Gender, male/female 34/10 Gestational weight at birth, g 3364 ± 446 (2560–4270) Head circumference at birth, cm 34.49 ± 1.2 (32–37) Gestational age at birth, weeks 39.66 ± 1.2 (37–41.7) Gestational age at first MRI, weeks 40.78 ± 2.18 (37.6–51.6) Time between first and second MRI scan, days 18.48 ± 6.81 (7–38) Socioeconomic status 9.09 ± 2.1 (5–12) Days of hospitalization 32.68 ± 11.43 (13–73) Bayley CCS 107.73 ± 14.32 (60–130) Bayley LCS 93.49 ± 14.076 (65–132) Bayley MCS 93.77 ± 13.51 (46–130) Continuously distributed data are displayed as mean ± SD (range). View Large Table 1 Demographic characteristics of the study population Total population Gender, male/female 34/10 Gestational weight at birth, g 3364 ± 446 (2560–4270) Head circumference at birth, cm 34.49 ± 1.2 (32–37) Gestational age at birth, weeks 39.66 ± 1.2 (37–41.7) Gestational age at first MRI, weeks 40.78 ± 2.18 (37.6–51.6) Time between first and second MRI scan, days 18.48 ± 6.81 (7–38) Socioeconomic status 9.09 ± 2.1 (5–12) Days of hospitalization 32.68 ± 11.43 (13–73) Bayley CCS 107.73 ± 14.32 (60–130) Bayley LCS 93.49 ± 14.076 (65–132) Bayley MCS 93.77 ± 13.51 (46–130) Total population Gender, male/female 34/10 Gestational weight at birth, g 3364 ± 446 (2560–4270) Head circumference at birth, cm 34.49 ± 1.2 (32–37) Gestational age at birth, weeks 39.66 ± 1.2 (37–41.7) Gestational age at first MRI, weeks 40.78 ± 2.18 (37.6–51.6) Time between first and second MRI scan, days 18.48 ± 6.81 (7–38) Socioeconomic status 9.09 ± 2.1 (5–12) Days of hospitalization 32.68 ± 11.43 (13–73) Bayley CCS 107.73 ± 14.32 (60–130) Bayley LCS 93.49 ± 14.076 (65–132) Bayley MCS 93.77 ± 13.51 (46–130) Continuously distributed data are displayed as mean ± SD (range). View Large The newborns underwent the following types of corrective surgeries: arterial switch or Rastelli operation for patients with dextro-transposition of the great arteries, Norwood-type stage I palliation for patients with hypoplastic left heart syndrome and other forms of univentricular disease with hypoplastic aortic arch, and complex aortic arch reconstruction for patients with interrupted aortic arch or coarctation of the aorta/transverse aortic arch hypoplasia as well as corrective cardiac surgery. The parents gave informed consent and the study was approved by the local ethical committee. The research was conducted according to the principles expressed in the Declaration of Helsinki. MRI acquisition protocols Neonatal cerebral MRI was performed in natural sleep on a 3.0 T clinical MRI scanner using an 8-channel head coil, and fast spin-echo T2-weighted anatomical sequences were acquired in axial, sagittal and coronal planes. The sequence parameters for the anatomical MRI were the following: echo time/repetition time: 97/5900 ms, flip angle: 90, pixel spacing: 0.35 × 0.35 mm, slice thickness: 2.7 mm. During the recruitment, the MRI scanner’s software was upgraded. Twenty-five cases were scanned before upgrade, but all pre-and postoperative scan pairs were performed on the same scanner software level. Newborns underwent the first, preoperative MRI at a corrected gestational age of 40.77 ± 2.18 (37.6–51.6) weeks, and the postoperative follow-up was performed at 43.52 ± 2.33 (40–54) weeks. The time difference between the two MRI exams were 18.48 ± 6.81 (range: 7–38) days. Image processing Figure 1 gives an overview of the image analysis steps used in the study. Magnetic resonance images were first masked using a semi-automated approach in the Slicer 3D software to remove voxels corresponding to non-brain tissue. The axial, sagittal and coronal T2-weighted images were reconstructed to a 3D image with isotropic voxel spacing of 1 mm using the BTK toolkit (Rousseau et al., 2013). Figure 1 View largeDownload slide Overview of the image acquisition and analysis steps. Figure 1 View largeDownload slide Overview of the image acquisition and analysis steps. To capture longitudinal changes, we used a modified version of deformation-based morphometry, in which the anatomical deformation that matches the pre- and postoperative images was calculated on the case level, and the resulting deformation field was spatially standardized across the population. The Jacobian determinant of a non-linear transformation field has previously been shown to feasibly capture longitudinal changes during development (Baloch et al., 2009; Hua et al., 2009), or in progressive brain volume loss (Cole et al., 2018). This was carried out using the following steps: (i) rigid alignment with 6 degrees of freedom of the preoperative, brain-extracted T2-weighted image with the postoperative MRI; (ii) calculation of the deformation field that matched the pre- and postoperative MRI based on a high-resolution free-form non-linear deformation algorithm (implemented in the NIFTIREG tool (Modat et al., 2010), reg_f3d command, control grid size: 5 mm isotropic, weight of the bending energy penalty term: 0.005, gradient smoothing with a kernel of 4 mm); (iii) transforming the natural logarithm of the Jacobian determinant of this transformation to brain growth rate (BGR) using the following equation: BGRi,j,k = lnJi,j,k(Tpostop − Tpreop) × lnJwb (1) where Ji,j,k ?> is the Jacobian determinant of the pre- to postoperative deformation field at voxel i,j,k, Jwb  ?> is the average Jacobian determinant over the brain tissue, (Tpostop−Tpreop) ?> is the time difference between the post- and preoperatively acquired magnetic resonance images. Because the times between the first and second MRI were highly variable, we calculated brain growth rate as the mean, daily change of the Jacobian determinant by dividing the growth by the time between the MRI scans. We then (iv) co-registered the preoperative neonatal brains to a standard, 40 gestational (term) week template image; (v) transformed the brain growth rate maps using the registration from step (iv) to the 40 gestational week template; and (vi) performed spatial smoothing with a Gaussian kernel of σ = 5 mm. Neurodevelopmental testing Neurodevelopmental testing was conducted using the Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III) at 12 months of age with three developmental domains: cognitive (CCS), language (LCS) and motor composite standard scores (MCS) with a mean score of 107.73, 93.5 and 93.77 in our cohort, respectively. Statistical tests Statistical analyses on the whole-brain brain growth rate maps were completed using the FSL software package (Jenkinson et al., 2012). Linear regression analysis was performed to evaluate the association of brain growth rate with the 1-year Bayley-III scores. We used the threshold-free cluster enhancement method (TFCE) (Smith and Nichols, 2009) with randomized non-parametric permutation testing implemented in the randomise program (Winkler et al., 2014), version 2.9, part of FSL (build 509) to correct for multiple comparisons. The default parameter settings of TFCE with 5000 permutations were used. The statistical analysis was restricted to a brain mask in template space corresponding to the brain parenchyma. For the initial analysis, family-wise error corrected P ≤ 0.025 was accepted as significant, this threshold accounted for the fact that positive and negative correlations were investigated separately. We chose to correct the first model for a variable that has been known to correlate with MRI appearance and hence may influence deformation-based morphometry results: ‘MRI software version’ (Shuter et al., 2008), since the scanner was upgraded during the recruitment period of the subjects. Furthermore, two known variables that may have influenced the neurocognitive tests were also included: parental socioeconomical status and gender (Largo et al., 1989; Wickremasinghe et al., 2012; Ronfani et al., 2015). The dichotomous variable ‘MRI software version’ was used to define two exchangeability blocks, and observations were permuted only within blocks. To demonstrate the effect size for linear regression analysis, maps depicting the Pearson product-moment correlation coefficient in MATLAB R2014 (Mathworks Inc., Mattick, USA) were calculated. Next, post hoc analysis was performed using stepwise linear regression in IBM SPSS V22 (IBM, Armonk, New York) to select any further demographic or clinical parameters that could affect the correlation between regional brain growth rate and the Bayley-III scores. During the stepwise selection of variables from a pool of 21 demographic and clinical parameters (Supplementary Table 2), the probability of F was set to 0.05 for a variable to enter and 0.1 for removal. Finally, a second linear regression analysis with randomized nonparametric permutation testing was performed, which was corrected for the confounding variables found in the post hoc analysis step, and TFCE-corrected P ≤ 0.05 was accepted as significant. Data availability The data that support the findings of this study are available on request from the corresponding author. The raw data are not publicly available due to the restrictions of using clinical records and images of the participants. Completely anonymized, secondary, i.e. processed neuroimaging datasets (brain growth rate maps) and the image processing scripts used in the study are available upon request from the corresponding author. Results Correlation between regional brain growth rate and 1-year Bayley-III composite scores We observed positive brain growth rate values over the entire brain parenchyma, reflecting increasing brain volumes during the perioperative period. Relative shrinkage was seen at the interface of the parenchyma and CSF spaces supratentorially, which was most likely the consequence of the mildly increasing ventricular dilatation over time. Whole-brain explorative analysis revealed that the brain growth rate of the left posterior perisylvian region was significantly positively correlated with the LCS at 12 months of age with no other clusters of correlation surviving the statistical significance threshold. This effect was localized to a circumscribed region comprising 1597 voxels (1.013 cm3) in one cluster, in which the peak TFCE-corrected significance (P = 0.0094, maximum t = 5.41, R = 0.557) was localized to the left temporal plane. The cluster extended laterally from the border of the left lateral ventricle towards the Heschl’s gyrus, respecting the grey matter and CSF interfaces along the anterior border of the Heschl’s gyrus (Fig. 2A), and covered anterior parts of the planum temporale. Superiorly, it extended into the white matter of the left parietal operculum. For brevity, this cluster is referred to hereafter as the temporal plane. A cluster of strong correlation between brain growth rate and LCS (R > 0.5) was co-localized mostly with the Heschl’s gyrus and in part with the anterior parts of the planum temporale on the correlation coefficient maps; however, additional clusters of moderate to strong, but not significant correlations (0.25 < R < 0.5) were also found around the frontal and occipital horns of the lateral ventricles and in the left frontal white matter (Fig. 2B). No significant positive correlation surviving the correction threshold of 0.025 was found between the brain growth rate and the CCS (P = 0.0475, corresponding to the maximum of t = 4.15) or MCS (P = 0.433, corresponding to the maximum of t = 3.59). No significant correlations were found between the brain growth rate maps and the CCS (P = 0.13, corresponding to the maximum of t = 3.77), the MCS (P = 0.64, corresponding to the maximum of t = 3.49) or the LCS (P = 0.28, corresponding to the maximum of t = 3.61). By qualitative inspection of weak to moderate correlations, we found that the correlation coefficient map of the CCS resembled greatly that of the LCS (Fig. 3), and the correlation between CCS and LCS was strong (R = 0.7249). CCS correlated more pronouncedly with brain growth rate in midline regions in the thalamus and bilaterally in the pre- and postcentral gyri. The correlation between MCS and brain growth rate showed increased bilaterality and left frontal dominance compared to LCS, and weak correlation was found between LCS and MCS (R = 0.1096). Contrary to our a priori hypothesis, brain growth rate did not correlate with socioeconomic status, gender or MRI software version when controlling for LCS. Figure 2 View largeDownload slide Correlation of perioperative regional brain growth rate with the 1-year Bayley-III Composite Language Score at 12 months of age. (A) Brain regions significantly (TFCE-corrected P ≤ 0.025) correlating with LCS are shown as red-to-yellow clusters overlaid on a T2-weighted MRI template, blue outline (within the red overlay): significant cluster after adjusting for length of hospital stay. (B) The correlation coefficient map (thresholded at R ≥ 0.25). (C) Correlation plot. ACM = two patients with middle cerebral artery stroke; ACP = one patient with posterior cerebral artery stroke; IVH = one patient with intraventricular haemorrhage reported on the preoperative MRI; PMCC = Pearson’s product-moment correlation coefficient. Figure 2 View largeDownload slide Correlation of perioperative regional brain growth rate with the 1-year Bayley-III Composite Language Score at 12 months of age. (A) Brain regions significantly (TFCE-corrected P ≤ 0.025) correlating with LCS are shown as red-to-yellow clusters overlaid on a T2-weighted MRI template, blue outline (within the red overlay): significant cluster after adjusting for length of hospital stay. (B) The correlation coefficient map (thresholded at R ≥ 0.25). (C) Correlation plot. ACM = two patients with middle cerebral artery stroke; ACP = one patient with posterior cerebral artery stroke; IVH = one patient with intraventricular haemorrhage reported on the preoperative MRI; PMCC = Pearson’s product-moment correlation coefficient. Figure 3 View largeDownload slide Comparison of brain growth rate correlating with the 1-year Bayley-III cognitive, motor and language scores. Brain regions with moderate to strong (R ≥ 0.25) positive correlation appear as a colour-coded overlay fused with the MRI template. PMCC = Pearson’s product-moment correlation coefficient. Figure 3 View largeDownload slide Comparison of brain growth rate correlating with the 1-year Bayley-III cognitive, motor and language scores. Brain regions with moderate to strong (R ≥ 0.25) positive correlation appear as a colour-coded overlay fused with the MRI template. PMCC = Pearson’s product-moment correlation coefficient. Post hoc analysis of disease severity, clinical and demographic effects We calculated the mean brain growth rate in the cluster in the temporal plane acquired from the previous analysis step, and post hoc stepwise linear regression analysis was performed to automatically select the possible clinical and demographic variables that may predict brain growth rate in the temporal plane cluster in combination. Two variables emerged from the stepwise linear regression, the LCS and the days of hospital stay (Table 2). The effect of left temporal plane growth on LCS was stronger with the length of hospital stay as covariate in the model, which means that the latter was a partial mediator of the main effect. While socioeconomic status positively correlated with LCS, it was not selected as a variable during the stepwise linear regression. Body weight, head circumference was not associated in any of the investigated variables (Table 3). Table 2 Post hoc stepwise linear regression analysis of the correlations between brain growth rate in the left temporal plane, LCS and significant confounders Variable Included in a priori test? Test statistic (t) Significance (P) Standardized coefficient (beta) Unstandardized coefficient (B, CI) LCSa Yes 5.198 6 × 10−6 0.626 0.027 (0.017–0.038) LCSb Yes 5.11 8 × 10−7 0.575 0.025 (0.015–0.035) Length of hospital stayb No 2.89 0.0061 0.325 0.017 (0.005–0.029) Variable Included in a priori test? Test statistic (t) Significance (P) Standardized coefficient (beta) Unstandardized coefficient (B, CI) LCSa Yes 5.198 6 × 10−6 0.626 0.027 (0.017–0.038) LCSb Yes 5.11 8 × 10−7 0.575 0.025 (0.015–0.035) Length of hospital stayb No 2.89 0.0061 0.325 0.017 (0.005–0.029) aRegression model, step 1. Adjusted R2 = 0.377. Model includes constant (intercept, not displayed). bRegression model, step 2. Adjusted R2 = 0.470. Model includes constant (intercept, not displayed). CI = confidence interval. View Large Table 2 Post hoc stepwise linear regression analysis of the correlations between brain growth rate in the left temporal plane, LCS and significant confounders Variable Included in a priori test? Test statistic (t) Significance (P) Standardized coefficient (beta) Unstandardized coefficient (B, CI) LCSa Yes 5.198 6 × 10−6 0.626 0.027 (0.017–0.038) LCSb Yes 5.11 8 × 10−7 0.575 0.025 (0.015–0.035) Length of hospital stayb No 2.89 0.0061 0.325 0.017 (0.005–0.029) Variable Included in a priori test? Test statistic (t) Significance (P) Standardized coefficient (beta) Unstandardized coefficient (B, CI) LCSa Yes 5.198 6 × 10−6 0.626 0.027 (0.017–0.038) LCSb Yes 5.11 8 × 10−7 0.575 0.025 (0.015–0.035) Length of hospital stayb No 2.89 0.0061 0.325 0.017 (0.005–0.029) aRegression model, step 1. Adjusted R2 = 0.377. Model includes constant (intercept, not displayed). bRegression model, step 2. Adjusted R2 = 0.470. Model includes constant (intercept, not displayed). CI = confidence interval. View Large Table 3 Pearson correlations among left temporal plane growth, demographic and neurodevelopmental outcome variables CCS LCS MCS Days of hospitalization Growth, left PT Body weight Head circumference LCS 0.578** – – – – – – MCS 0.596** 0.420** – – – – – Days of hospitalization −0.078 0.157 −0.208 – – – – Growth, left TP 0.539** 0.635** 0.323* 0.419** – – – Body weight 0.168 0.020 0.250 −0.191 0.067 – – Head circumference −0.032 0.206 −0.081 0.142 0.188 0.018 – SES 0.214 0.308* 0.162 −0.057 0.011 0.104 0.012 CCS LCS MCS Days of hospitalization Growth, left PT Body weight Head circumference LCS 0.578** – – – – – – MCS 0.596** 0.420** – – – – – Days of hospitalization −0.078 0.157 −0.208 – – – – Growth, left TP 0.539** 0.635** 0.323* 0.419** – – – Body weight 0.168 0.020 0.250 −0.191 0.067 – – Head circumference −0.032 0.206 −0.081 0.142 0.188 0.018 – SES 0.214 0.308* 0.162 −0.057 0.011 0.104 0.012 *Correlation is significant at the 0.05 level (2-tailed). **Correlation is significant at the 0.01 level (2-tailed). PT = planum temporale. View Large Table 3 Pearson correlations among left temporal plane growth, demographic and neurodevelopmental outcome variables CCS LCS MCS Days of hospitalization Growth, left PT Body weight Head circumference LCS 0.578** – – – – – – MCS 0.596** 0.420** – – – – – Days of hospitalization −0.078 0.157 −0.208 – – – – Growth, left TP 0.539** 0.635** 0.323* 0.419** – – – Body weight 0.168 0.020 0.250 −0.191 0.067 – – Head circumference −0.032 0.206 −0.081 0.142 0.188 0.018 – SES 0.214 0.308* 0.162 −0.057 0.011 0.104 0.012 CCS LCS MCS Days of hospitalization Growth, left PT Body weight Head circumference LCS 0.578** – – – – – – MCS 0.596** 0.420** – – – – – Days of hospitalization −0.078 0.157 −0.208 – – – – Growth, left TP 0.539** 0.635** 0.323* 0.419** – – – Body weight 0.168 0.020 0.250 −0.191 0.067 – – Head circumference −0.032 0.206 −0.081 0.142 0.188 0.018 – SES 0.214 0.308* 0.162 −0.057 0.011 0.104 0.012 *Correlation is significant at the 0.05 level (2-tailed). **Correlation is significant at the 0.01 level (2-tailed). PT = planum temporale. View Large Next, the effect of the length of hospital stay on the brain growth rate over the whole brain was evaluated, controlling for LCS, socioeconomic status, gender and MRI software version (Fig. 4). We found that the length of hospital stay positively correlated with brain growth rate independently from LCS in both hemispheres periventricularly, indicating excessive growth of CSF spaces and the central parts of the white matter, the volume of the significant cluster was 8565 voxels (5.43 cm3) with a peak of P = 0.018 and t = 5.656 (Fig. 4A). Local volume reduction was associated with length of hospital stay over the entire brain surface; however, significance was not reached after TFCE correction (Fig. 4B). No further parameters related to disease severity contributed significantly to this model. Results of the linear regression for variables not selected are displayed in Supplementary Table 2. Figure 4 View largeDownload slide Effect of length of hospital stay on regional brain growth rate, corrected for LCS, SES, gender and MRI software version. (A) Test statistic maps (t), positive values referring to positive correlation between regional growth rate and hospitalization duration. (B) Brain regions with significant (TFCE-corrected P ≤ 0.025) positive correlation are depicted as red-to-yellow overlays, and were fused with the MRI template. The brain region in which brain growth rate correlated with LCS is displayed as white outlines. Figure 4 View largeDownload slide Effect of length of hospital stay on regional brain growth rate, corrected for LCS, SES, gender and MRI software version. (A) Test statistic maps (t), positive values referring to positive correlation between regional growth rate and hospitalization duration. (B) Brain regions with significant (TFCE-corrected P ≤ 0.025) positive correlation are depicted as red-to-yellow overlays, and were fused with the MRI template. The brain region in which brain growth rate correlated with LCS is displayed as white outlines. Three subjects had ischaemic lesions (two cases with arteria cerebri media and one case with arteria cerebri posterior stroke, see the Supplementary material) and one intra-ventricular hemorrhage, the latter was detected on the preoperative MRI. Because of the small number of such cases, we decided not to include a variable for ischaemic lesions as a confounder but demonstrated the relationship between brain growth rate and cognitive outcomes in these cases individually (Fig. 2C). The LCS and the temporal plane cluster’s growth rate were markedly lower in the three cases with ischaemic lesions, two of them demonstrated a shrinkage of the temporal plane cluster in the postoperative MRI. The one case with intra-ventricular haemorrhage demonstrated average regional brain growth rate in the temporal plane and a normal LCS score (109) at 12 months. To rule out the contribution of the possible anatomical abnormalities linked to perinatal stroke, we additionally tested the main hypothesis test excluding these three cases. Due to the fact that the overall effect of left temporal plane brain growth on the variability on LCS did not change, we only report these findings in the Supplementary material. Correlation between regional brain growth and 1-year Bayley-III scores, controlling for the length of hospital stay A second linear regression analysis with randomized non-parametric permutation testing was performed, controlling for the length of hospital stay, socioeconomic status, gender and MRI software version. brain growth rate correlated positively with LCS after TFCE correction (P ≤ 0.05) in the left hemispheric perisylvian region, similar to the model not including the length of hospital stay as a confounder. This well-circumscribed cluster comprised 861 voxels (0.546 cm3), peak TFCE-corrected significance was P = 0.0268, maximum t = 4.4 (shown in Fig. 2A as blue outlines). Discussion CHD infants demonstrated global and regional cerebral growth within the first 3 weeks after cardiopulmonary bypass surgery and intensive care stay. While regional brain growth rates during this period showed no significant association with motor or general cognitive outcomes, perioperative growth in the left-dominant perisylvian region was strongly correlated with language performance at 1 year of age. This close relationship was observed during the critical life period of CHD infants when they underwent the first surgery for the underlying abnormality. In a multivariate model including gender, MRI software version and socioeconomic status, regional brain growth rate of the left Heschl’s gyrus and anterior parts of planum temporale and parietal operculum correlated strongly with language development at 12 months of age. Whereas this newly described morphological substrate for early language development appeared to be left lateralized and well-circumscribed, such high anatomical specificity was not observed for the cognitive and motor development of the infants. The regions in which regional growth moderately correlated with CCS comprised the bilateral pre- and postcentral gyri, the medial thalamus, posterior, the medial occipital white matter and the posterior horn of the lateral ventricle. This may reflect that CCS and MCS scores are more generalized markers of neurodevelopment, with more distributed neuroanatomical substrates and less pronounced morphology-function link at such an early age. The length of hospital stay correlated with brain growth rate around the CSF spaces, which could reflect increasingly dilated lateral ventricles or volume loss of the cerebral white matter surrounding the ventricles. We showed that the length of hospital stay was a partial mediator of the correlation between brain growth rate and neurodevelopmental outcome. A possible explanation for this is that hospitalization reflected postoperative disease severity. It could be affected by low cardiac output syndrome including low cerebral perfusion, longer need for inotropic support and longer time of intubation. Other factors, such as infections, rhythm disorders, renal insufficiency and higher nutritive demands may affect cerebral growth. Dilatation of the lateral ventricles has been shown in older children at 2–3 years of age before Fontan procedure to correlate with adverse neurodevelopmental outcome (Knirsch et al., 2016). Our results are in agreement with previous works that found a link between left perisylvian morphology and language development in high-risk infants, such as in newborns suffering from hypoxic-ischaemic injury (Shapiro et al., 2017) or preterm infants (Aeby et al., 2013). In CHD, white matter volume was found to correlate with language development but not broader developmental indices (Rollins et al., 2017), similar to our measurements, and in concordance with studies on extremely preterm infants (Skiold et al., 2014). In our study, the strongest correlation with language development was observed in the left Heschl’s gyrus, a region responsible for processing incoming auditory information as part of the primary auditory cortex. Impaired microstructural maturation of the auditory cortex was recently described to be predictive of poorer language performance in preterm infants (Monson et al., 2018), most likely because of the increased vulnerability of regions responsible for higher level sensory processing. Reduced hemispheric asymmetry patterns in the temporal lobe in preterm infants were found to be linked to reduced language performance (Pineda et al., 2014). Failure to develop left-hemispheric specialization may therefore be the sign of maladaptive neurodevelopment, which could be the consequence of suboptimal sensory stimulation during hospitalization (Pineda et al., 2014). By the same token, infants with CHD spend extensive periods in the neonatal intensive care unit environment, which may adversely affect early auditory development. In fact, this could be an alternative explanation why the length of hospitalization was a partial mediator of the main correlation in our study, implying possible similarities between the pathogenesis of impaired auditory development in CHD and preterm infants. Impaired development of the auditory system during the critical perioperative period may also in part be responsible for the very high prevalence of hearing loss in preschool children after heart surgery in infancy (Grasty et al., 2018). While previous studies reported links between morphological features at an early stage of neurodevelopment and later life cognitive impairment (Miller et al., 2007; Licht et al., 2009; Aeby et al., 2013; Skiold et al., 2014; Rollins et al., 2017; Shapiro et al., 2017), it is not yet clear if these findings could directly be attributed to the underlying pathology or rather reflect interindividual variability in normal development. Our study sample consisted of infants with subnormal and normal cognitive scores and therefore represents a population with impaired development rather than the normal variability. We demonstrated that the brain growth rate in the left temporal plane over the first 3 weeks of life is predictive of later language performance in CHD. This marks a high anatomical specificity of language processing immediately after birth, providing further evidence that language processing in infants is supported by neuroanatomical structures similar to the adult brain early on. The primary auditory cortex is able to process inputs during the second trimester of gestation (Krmpotic-Nemanic et al., 1983); however, the non-primary auditory cortex continues to mature afterwards (Monson et al., 2018). Based on animal models, it is safe to assume that adverse inputs, such as noisy environment, may interfere with the later life refinement of cortical functional areas (Chang and Merzenich, 2003). The left superior temporal and angular gyri are already active in infants during speech processing at 3 months of age (Dehaene-Lambertz et al., 2002; Pena et al., 2003; Dehaene-Lambertz et al., 2006; Arimitsu et al., 2011). These functional neuroimaging studies suggest that language processing is already localized to the left perisylvian region shortly after or even before birth. The pre-existing hemispheric asymmetry of the perisylvian region most likely underlies the functional specialization for language at such an early age: it is known that neonates are born with left dominant planum temporale and deeper right superior temporal sulcus (Witelson and Pallie, 1973; Kasprian et al., 2011; Schuler et al., 2017), and the development of language structures supporting processing areas also demonstrate hemispheric asymmetry in newborns (Dubois et al., 2008; Dubois et al., 2010, 2014; Ratnarajah et al., 2013). These in utero and early postnatal findings imply strong genetic influence on the observed perisylvian asymmetry, which is further shaped by exposure to language, resulting in more pronounced lateralization patterns seen in later life (Rosselli et al., 2014; Kwon et al., 2015). Similarly, pathological processes can interfere with language development during infancy, leading to neuroplastic reorganization of language-processing areas in cases of structural damage to the left perisylvian region (Dehaene-Lambertz et al., 2004). Reduced left-hemispheric specialization indicates maladaptive neurodevelopment, such as in autism spectrum disorders (Eyler et al., 2012). While surgical correction and palliation of CHD improved survival rates in CHD, the lag behind normal neurodevelopment appears to persist despite corrective surgery (Latal, 2016). The impairment of higher cognitive functions becomes most apparent during childhood or adolescence, hallmarking the importance of close neurodevelopmental follow-up. Single time point studies confirmed a variety of neuroanatomical correlates of cognitive development in CHD, such as reduced global brain volumes (Miller et al., 2007; Limperopoulos et al., 2010; Ortinau et al., 2012a, b; von Rhein et al., 2014, 2015), cerebral white matter volume and microstructure (Rollins et al., 2014, 2017), and hippocampal volume reduction (Latal et al., 2016). There is also a lack of evidence whether individual differences in early pre- or postoperative growth rates are useful predictors of cognitive outcome in complex CHD. Longitudinally acquired cerebral MRI offers the possibility to non-invasively characterize the trajectories of structural brain development in children with complex CHD and to find a link between growth and long-term cognitive development. Previous research found that cerebral growth rate following surgical correction may not be different in the first 3 months of life in a mixed group of complex CHD compared to typical development (Ortinau et al., 2012a). On the contrary, it was shown that growth trajectories might exhibit differences depending on the underlying physiology (Peyvandi et al., 2018). Correlation between growth rates and neurodevelopment was confirmed for other risk groups and pathologies. In preterms, somatic growth velocity during hospitalization was proven to be a possible independent predictor of adverse neurodevelopment (Ehrenkranz et al., 2006). Impaired brain growth trajectories were associated with foetal alcohol spectrum disorders, a condition also characterized by cognitive impairment (Treit et al., 2013; Maruyama et al., 2015). Neurocognitive outcomes in high risk infants may, however, be better predicted by the multi-parametric analysis of brain morphology, brain connectivity and function (Ball et al., 2015; Paquette et al., 2015). The following limitations of our study merit mentioning. We observed a remarkably large effect of brain growth over a rather short period of time—on average 3 weeks—on language outcomes. This relatively short time and the high variability of the time between the two MRI scans were limitations in our study design. It is known that the maturation of the white matter during the first months of life is particularly rapid in cortico-thalamic, corticospinal fibres, as well as in subcortical areas (Welker and Patton, 2012), and that the morphological maturation of brain convolutions continue after birth. Based on this knowledge, we presume that variability of the left temporal plane growth contributed significantly to the differences seen in the language development; however, other factors, such as perioperative brain injury, the surgical technique and ischaemic lesions or intraventricular haemorrhage might also have influenced the observed relationship between growth and outcome. We found no relationship between the presence of T2-hyperintensity, white matter damage and left temporal plane growth; however, we were not able to discern such link for ischaemic brain lesions because of the low number of such cases (3/44). The heterogeneity in our sample calls for careful attention when generalizing the results to all types of CHD, and long-term follow-up results will be necessary to validate if the revealed morphology-outcome link is sustained into later stages of development. It remains an open question if a similar correlation between early neonatal brain growth in the left temporal plane and 1-year language outcome exists in other risk groups, such as in preterm or term born, normally developing infants. The assessment of language impairments at 1 year is limited as language development at that age is not yet complete, and future works are needed to replicate our findings by using more detailed and later assessments of language development. Our findings are in line with Rollins et al. (2017), who described that reduced white matter volume correlates positively with MacArthur-Bates Communicative Development Inventory language development. On the contrary, Peyvandi et al. (2018) showed that, indeed, later motor performance at 30 months was better predicted by perioperative white matter injury compared to functional performance at 12 months, which emphasizes the fact that long term development might not reliably be predicted with the tests that our study used at 1 year of age. Our study also lacks a comprehensive evaluation of auditory development, which would be necessary in the future to prove the link between impaired brain maturation in the perioperative period and the high prevalence of hearing loss in CHD patients (Grasty et al., 2018). While the lack of controls hinders the more general interpretation of our results, we hypothesize that CHD remains an optimal model to study such correlations because of the higher variability of cognitive outcomes and growth rates in this population. The interpretation of our results is further limited by the fact that the deformation-based morphometry method could not distinguish between white matter and cortical growth, as the images were not segmented into such compartments prior to analysis. Despite this, deformation-based morphometry might be appropriate for the analysis of the newborn brain, as brain maturation results in rapid changes in the MRI appearance, rendering automatic tissue type classification difficult (Aljabar et al., 2008). In our study, we did not determine brain weight (corrected for body weight) as a result parameter in our study because of the specific analysis of very small regional parts of the brain. Our study provides further evidence for the early importance of left-dominant perisylvian regions in language development before the lasting exposure to native language. Interindividual variability of regional brain growth rates in this area was found to be linked to variability in language performance at 12 months of age. Deformation-based morphometry, based on longitudinal MRI before and after the corrective surgery for CHD, may help to identify at-risk infants for impaired auditory and language development, and enable better clinical counselling for high risk infants. Funding A.J. was supported by the OPO Foundation, the FZK Foundation, the Stiftung für wissenschaftliche Forschung an der UZH and the EMDO Foundation, R.T. was supported by the EMDO Foundation, B.L. was supported by the Mäxi Foundation. Competing interests The authors report no competing interests. Appendix 1 The Research Group Heart and Brain includes the following members who are not listed as authors: Ingrid Beck, Vera Bernet, Cornelia Hagmann, Michael Huebler, Oliver Kretschmar, Rabia Liamlahi, Annette Hackenberg. Abbreviations Abbreviations CCS Bayley-III Cognitive Composite Score CHD congenital heart disease MCS Bayley-III Motor Composite Score LCS Bayley-III Language Composite Score TFCE threshold-free cluster enhancement References Aeby A , De Tiege X , Creuzil M , David P , Baleriaux D , Van Overmeire B et al. . Language development at 2 years is correlated to brain microstructure in the left superior temporal gyrus at term equivalent age: a diffusion tensor imaging study . Neuroimage 2013 ; 78 : 145 – 51 . Google Scholar Crossref Search ADS PubMed Aljabar P , Bhatia KK , Murgasova M , Hajnal JV , Boardman JP , Srinivasan L et al. . Assessment of brain growth in early childhood using deformation-based morphometry . Neuroimage 2008 ; 39 : 348 – 58 . Google Scholar Crossref Search ADS PubMed Andropoulos DB , Ahmad HB , Haq T , Brady K , Stayer SA , Meador MR et al. . The association between brain injury, perioperative anesthetic exposure, and 12-month neurodevelopmental outcomes after neonatal cardiac surgery: a retrospective cohort study . Paediatr Anaesth 2014 ; 24 : 266 – 74 . Google Scholar Crossref Search ADS PubMed Arimitsu T , Uchida-Ota M , Yagihashi T , Kojima S , Watanabe S , Hokuto I et al. . Functional hemispheric specialization in processing phonemic and prosodic auditory changes in neonates . Front Psychol 2011 ; 2 : 202 . Google Scholar Crossref Search ADS PubMed Ball G , Pazderova L , Chew A , Tusor N , Merchant N , Arichi T et al. . Thalamocortical connectivity predicts cognition in children born preterm . Cereb Cortex 2015 ; 25 : 4310 – 8 . Google Scholar Crossref Search ADS PubMed Baloch S , Verma R , Huang H , Khurd P , Clark S , Yarowsky P et al. . Quantification of brain maturation and growth patterns in C57BL/6J mice via computational neuroanatomy of diffusion tensor images . Cereb Cortex 2009 ; 19 : 675 – 87 . Google Scholar Crossref Search ADS PubMed Bellinger DC , Newburger JW , Wypij D , Kuban KC , duPlesssis AJ , Rappaport LA . Behaviour at eight years in children with surgically corrected transposition: the Boston Circulatory Arrest Trial . Cardiol Young 2009 ; 19 : 86 – 97 . Google Scholar Crossref Search ADS PubMed Bellinger DC , Wypij D , Kuban KC , Rappaport LA , Hickey PR , Wernovsky G et al. . Developmental and neurological status of children at 4 years of age after heart surgery with hypothermic circulatory arrest or low-flow cardiopulmonary bypass . Circulation 1999 ; 100 : 526 – 32 . Google Scholar Crossref Search ADS PubMed Cassidy AR , White MT , DeMaso DR , Newburger JW , Bellinger DC . Executive function in children and adolescents with critical cyanotic congenital heart disease . J Int Neuropsychol Soc 2015 ; 21 : 34 – 49 . Google Scholar Crossref Search ADS PubMed Chang EF , Merzenich MM . Environmental noise retards auditory cortical development . Science 2003 ; 300 : 498 – 502 . Google Scholar Crossref Search ADS PubMed Claessens NH , Moeskops P , Buchmann A , Latal B , Knirsch W , Scheer I et al. . Delayed cortical gray matter development in neonates with severe congenital heart disease . Pediatr Res 2016 ; 80 : 668 – 74 . Google Scholar Crossref Search ADS PubMed Cole JH , Jolly A , de Simoni S , Bourke N , Patel MC , Scott G et al. . Spatial patterns of progressive brain volume loss after moderate-severe traumatic brain injury . Brain 2018 ; 141 : 822 – 36 . Google Scholar Crossref Search ADS PubMed Dehaene-Lambertz G , Dehaene S , Hertz-Pannier L . Functional neuroimaging of speech perception in infants . Science 2002 ; 298 : 2013 – 5 . Google Scholar Crossref Search ADS PubMed Dehaene-Lambertz G , Pena M , Christophe A , Landrieu P . Phoneme perception in a neonate with a left sylvian infarct . Brain Lang 2004 ; 88 : 26 – 38 . Google Scholar Crossref Search ADS PubMed Dehaene-Lambertz G , Hertz-Pannier L , Dubois J , Meriaux S , Roche A , Sigman M et al. . Functional organization of perisylvian activation during presentation of sentences in preverbal infants . Proc Natl Acad Sci USA 2006 ; 103 : 14240 – 5 . Google Scholar Crossref Search ADS PubMed Dubois J , Dehaene-Lambertz G , Kulikova S , Poupon C , Hüppi PS , Hertz-Pannier L . The early development of brain white matter: a review of imaging studies in fetuses, newborns and infants . Neuroscience 2014 ; 276 : 48 – 71 . Google Scholar Crossref Search ADS PubMed Dubois J , Benders M , Lazeyras F , Borradori-Tolsa C , Leuchter RH , Mangin JF et al. . Structural asymmetries of perisylvian regions in the preterm newborn . Neuroimage 2010 ; 52 : 32 – 42 . Google Scholar Crossref Search ADS PubMed Dubois J , Dehaene-Lambertz G , Perrin M , Mangin JF , Cointepas Y , Duchesnay E et al. . Asynchrony of the early maturation of white matter bundles in healthy infants: quantitative landmarks revealed noninvasively by diffusion tensor imaging . Hum Brain Mapp 2008 ; 29 : 14 – 27 . Google Scholar Crossref Search ADS PubMed Ehrenkranz RA , Dusick AM , Vohr BR , Wright LL , Wrage LA , Poole WK . Growth in the neonatal intensive care unit influences neurodevelopmental and growth outcomes of extremely low birth weight infants . Pediatrics 2006 ; 117 : 1253 – 61 . Google Scholar Crossref Search ADS PubMed Eyler LT , Pierce K , Courchesne E . A failure of left temporal cortex to specialize for language is an early emerging and fundamental property of autism . Brain 2012 ; 135 : 949 – 60 . Google Scholar Crossref Search ADS PubMed Grasty MA , Ittenbach RF , Knightly C , Solot CB , Gerdes M , Bernbaum JC et al. . Hearing Loss after cardiac surgery in infancy: an unintended consequence of life-saving care . J Pediatr 2018 ; 192 : 144, 151.e1 . Google Scholar Crossref Search ADS Hovels-Gurich HH , Bauer SB , Schnitker R , Willmes-von Hinckeldey K , Messmer BJ , Seghaye MC et al. . Long-term outcome of speech and language in children after corrective surgery for cyanotic or acyanotic cardiac defects in infancy . Eur J Paediatr Neurol 2008 ; 12 : 378 – 86 . Google Scholar Crossref Search ADS PubMed Hovels-Gurich HH , Konrad K , Skorzenski D , Nacken C , Minkenberg R , Messmer BJ et al. . Long-term neurodevelopmental outcome and exercise capacity after corrective surgery for tetralogy of Fallot or ventricular septal defect in infancy . Ann Thorac Surg 2006 ; 81 : 958 – 66 . Google Scholar Crossref Search ADS PubMed Hovels-Gurich HH , Seghaye MC , Schnitker R , Wiesner M , Huber W , Minkenberg R et al. . Long-term neurodevelopmental outcomes in school-aged children after neonatal arterial switch operation . J Thorac Cardiovasc Surg 2002 ; 124 : 448 – 58 . Google Scholar Crossref Search ADS PubMed Hua X , Leow AD , Levitt JG , Caplan R , Thompson PM , Toga AW . Detecting brain growth patterns in normal children using tensor-based morphometry . Hum Brain Mapp 2009 ; 30 : 209 – 19 . Google Scholar Crossref Search ADS PubMed Jenkinson M , Beckmann CF , Behrens TE , Woolrich MW , Smith SM . Fsl . Neuroimage 2012 ; 62 : 782 – 90 . Google Scholar Crossref Search ADS PubMed Kasprian G , Langs G , Brugger PC , Bittner M , Weber M , Arantes M et al. . The prenatal origin of hemispheric asymmetry: an in utero neuroimaging study . Cereb Cortex 2011 ; 21 : 1076 – 83 . Google Scholar Crossref Search ADS PubMed Kirshbom PM , Myung RJ , Gaynor JW , Ittenbach RF , Paridon SM , DeCampli WM et al. . Preoperative pulmonary venous obstruction affects long-term outcome for survivors of total anomalous pulmonary venous connection repair . Ann Thorac Surg 2002 ; 74 : 1616 – 20 . Google Scholar Crossref Search ADS PubMed Knirsch W , Liamlahi R , Dave H , Kretschmar O , Latal B . Neurodevelopmental outcome of children with hypoplastic left heart syndrome at one and four years of age comparing hybrid and norwood procedure . Ann Thorac Cardiovasc Surg 2016 ; 22 : 375 – 7 . Google Scholar Crossref Search ADS PubMed Krmpotic-Nemanic J , Kostovic I , Kelovic Z , Nemanic D , Mrzljak L . Development of the human fetal auditory cortex: growth of afferent fibres . Acta Anat (Basel) 1983 ; 116 : 69 – 73 . Google Scholar Crossref Search ADS PubMed Kwon SH , Scheinost D , Lacadie C , Sze G , Schneider KC , Dai F et al. . Adaptive mechanisms of developing brain: cerebral lateralization in the prematurely-born . Neuroimage 2015 ; 108 : 144 – 50 . Google Scholar Crossref Search ADS PubMed Largo RH , Pfister D , Molinari L , Kundu S , Lipp A , Duc G . Significance of prenatal, perinatal and postnatal factors in the development of AGA preterm infants at five to seven years . Dev Med Child Neurol 1989 ; 31 : 440 – 56 . Google Scholar Crossref Search ADS PubMed Latal B . Neurodevelopmental outcomes of the child with congenital heart disease . Clin Perinatol 2016 ; 43 : 173 – 85 . Google Scholar Crossref Search ADS PubMed Latal B , Patel P , Liamlahi R , Knirsch W , O’Gorman Tuura R , von Rhein M . Hippocampal volume reduction is associated with intellectual functions in adolescents with congenital heart disease . Pediatr Res 2016 ; 80 : 531 – 7 . Google Scholar Crossref Search ADS PubMed Licht DJ , Shera DM , Clancy RR , Wernovsky G , Montenegro LM , Nicolson SC et al. . Brain maturation is delayed in infants with complex congenital heart defects . J Thorac Cardiovasc Surg 2009 ; 137 : 529 – 36; discussion 536–7 . Google Scholar Crossref Search ADS PubMed Limperopoulos C , Majnemer A , Shevell MI , Rohlicek C , Rosenblatt B , Tchervenkov C et al. . Predictors of developmental disabilities after open heart surgery in young children with congenital heart defects . J Pediatr 2002 ; 141 : 51 – 8 . Google Scholar Crossref Search ADS PubMed Limperopoulos C , Tworetzky W , McElhinney DB , Newburger JW , Brown DW , Robertson RL Jr et al. . Brain volume and metabolism in fetuses with congenital heart disease: evaluation with quantitative magnetic resonance imaging and spectroscopy . Circulation 2010 ; 121 : 26 – 33 . Google Scholar Crossref Search ADS PubMed Maruyama H , Yonemoto N , Kono Y , Kusuda S , Fujimura M , Neonatal research network of Japan. Weight growth velocity and neurodevelopmental outcomes in extremely low birth weight infants . PLoS One 2015 ; 10 : e0139014 . Google Scholar Crossref Search ADS PubMed McQuillen PS , Goff DA , Licht DJ . Effects of congenital heart disease on brain development . Prog Pediatr Cardiol 2010 ; 29 : 79 – 85 . Google Scholar Crossref Search ADS PubMed Miller SP , McQuillen PS , Hamrick S , Xu D , Glidden DV , Charlton N et al. . Abnormal brain development in newborns with congenital heart disease . N Engl J Med 2007 ; 357 : 1928 – 38 . Google Scholar Crossref Search ADS PubMed Modat M , Ridgway GR , Taylor ZA , Lehmann M , Barnes J , Hawkes DJ et al. . Fast free-form deformation using graphics processing units . Comput Methods Programs Biomed 2010 ; 98 : 278 – 84 . Google Scholar Crossref Search ADS PubMed Monson BB , Eaton-Rosen Z , Kapur K , Liebenthal E , Brownell A , Smyser CD et al. . Differential rates of perinatal maturation of human primary and nonprimary auditory cortex . eNeuro 2018 ; 5 , doi: 10.1523/ENEURO.0380–17.2017. eCollection 2018 Jan-Feb . Naef N , Liamlahi R , Beck I , Bernet V , Dave H , Knirsch W et al. . Neurodevelopmental profiles of children with congenital heart disease at school age . J Pediatr 2017 ; 188 : 75 – 81 . Google Scholar Crossref Search ADS PubMed Ortinau C , Inder T , Lambeth J , Wallendorf M , Finucane K , Beca J . Congenital heart disease affects cerebral size but not brain growth . Pediatr Cardiol 2012a ; 33 : 1138 – 46 . Google Scholar Crossref Search ADS Ortinau C , Beca J , Lambeth J , Ferdman B , Alexopoulos D , Shimony JS et al. . Regional alterations in cerebral growth exist preoperatively in infants with congenital heart disease . J Thorac Cardiovasc Surg 2012b ; 143 : 1264 – 70 . Google Scholar Crossref Search ADS Owen M , Shevell M , Donofrio M , Majnemer A , McCarter R , Vezina G et al. . Brain volume and neurobehavior in newborns with complex congenital heart defects . J Pediatr 2014 ; 164 : 1121 – 27.e1 . Google Scholar Crossref Search ADS PubMed Paquette LB , Votava-Smith JK , Ceschin R , Nagasunder AC , Jackson HA , Bluml S et al. . Abnormal development of thalamic microstructure in premature neonates with congenital heart disease . Pediatr Cardiol 2015 ; 36 : 960 – 9 . Google Scholar Crossref Search ADS PubMed Pena M , Maki A , Kovacic D , Dehaene-Lambertz G , Koizumi H , Bouquet F et al. . Sounds and silence: an optical topography study of language recognition at birth . Proc Natl Acad Sci USA 2003 ; 100 : 11702 – 5 . Google Scholar Crossref Search ADS PubMed Peyvandi S , Kim H , Lau J , Barkovich AJ , Campbell A , Miller S et al. . The association between cardiac physiology, acquired brain injury, and postnatal brain growth in critical congenital heart disease . J Thorac Cardiovasc Surg 2018 ; 155 : 291, 300.e3 . Google Scholar Crossref Search ADS Pineda RG , Neil J , Dierker D , Smyser CD , Wallendorf M , Kidokoro H et al. . Alterations in brain structure and neurodevelopmental outcome in preterm infants hospitalized in different neonatal intensive care unit environments . J Pediatr 2014 ; 164 : 52 – 60.e2 . Google Scholar Crossref Search ADS PubMed Ratnarajah N , Rifkin-Graboi A , Fortier MV , Chong YS , Kwek K , Saw SM et al. . Structural connectivity asymmetry in the neonatal brain . Neuroimage 2013 ; 75 : 187 – 94 . Google Scholar Crossref Search ADS PubMed Rollins CK , Asaro LA , Akhondi-Asl A , Kussman BD , Rivkin MJ , Bellinger DC et al. . White matter volume predicts language development in congenital heart disease . J Pediatr 2017 ; 181 : 42 – 48.e2 . Google Scholar Crossref Search ADS PubMed Rollins CK , Watson CG , Asaro LA , Wypij D , Vajapeyam S , Bellinger DC et al. . White matter microstructure and cognition in adolescents with congenital heart disease . J Pediatr 2014 ; 165 : 936 – 44.e1–2 . Google Scholar Crossref Search ADS PubMed Ronfani L , Vecchi Brumatti L , Mariuz M , Tognin V , Bin M , Ferluga V et al. . The complex interaction between home environment, socioeconomic status, maternal IQ and early child neurocognitive development: a multivariate analysis of data collected in a Newborn Cohort Study . PLoS One 2015 ; 10 : e0127052 . Google Scholar Crossref Search ADS PubMed Rosselli M , Ardila A , Matute E , Velez-Uribe I . Language development across the life span: a neuropsychological/neuroimaging perspective . Neurosci J 2014 ; 2014 : 585237 . Google Scholar Crossref Search ADS PubMed Rousseau F , Oubel E , Pontabry J , Schweitzer M , Studholme C , Koob M et al. . BTK: an open-source toolkit for fetal brain MR image processing . Comput Methods Programs Biomed 2013 ; 109 : 65 – 73 . Google Scholar Crossref Search ADS PubMed Schaefer C , von Rhein M , Knirsch W , Huber R , Natalucci G , Caflisch J et al. . Neurodevelopmental outcome, psychological adjustment, and quality of life in adolescents with congenital heart disease . Dev Med Child Neurol 2013 ; 55 : 1143 – 9 . Google Scholar Crossref Search ADS PubMed Schuler AL , Kasprian G , Schwartz E , Seidl R , Diogo MC , Mitter C et al. . Mens inversus in corpore inverso? Language lateralization in a boy with situs inversus totalis . Brain Lang 2017 ; 174 : 9 – 15 . Google Scholar Crossref Search ADS PubMed Shapiro KA , Kim H , Mandelli ML , Rogers EE , Gano D , Ferriero DM et al. . Early changes in brain structure correlate with language outcomes in children with neonatal encephalopathy . Neuroimage Clin 2017 ; 15 : 572 – 80 . Google Scholar Crossref Search ADS PubMed Shuter B , Yeh IB , Graham S , Au C , Wang SC . Reproducibility of brain tissue volumes in longitudinal studies: effects of changes in signal-to-noise ratio and scanner software . Neuroimage 2008 ; 41 : 371 – 9 . Google Scholar Crossref Search ADS PubMed Skiold B , Alexandrou G , Padilla N , Blennow M , Vollmer B , Aden U . Sex differences in outcome and associations with neonatal brain morphology in extremely preterm children . J Pediatr 2014 ; 164 : 1012 – 8 . Google Scholar Crossref Search ADS PubMed Smith SM , Nichols TE . Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference . Neuroimage 2009 ; 44 : 83 – 98 . Google Scholar Crossref Search ADS PubMed Snookes SH , Gunn JK , Eldridge BJ , Donath SM , Hunt RW , Galea MP et al. . A systematic review of motor and cognitive outcomes after early surgery for congenital heart disease . Pediatrics 2010 ; 125 : e818 – 27 . Google Scholar Crossref Search ADS PubMed Treit S , Lebel C , Baugh L , Rasmussen C , Andrew G , Beaulieu C . Longitudinal MRI reveals altered trajectory of brain development during childhood and adolescence in fetal alcohol spectrum disorders . J Neurosci 2013 ; 33 : 10098 – 109 . Google Scholar Crossref Search ADS PubMed Turan S , Rosenbloom JI , Hussein M , Berg C , Gembruch U , Baschat AA et al. . Longitudinal analysis of head and somatic growth in fetuses with congenital heart defects . J Clin Ultrasound 2017 ; 45 : 96 – 104 . Google Scholar Crossref Search ADS PubMed Volpe JJ . Encephalopathy of congenital heart disease- destructive and developmental effects intertwined . J Pediatr 2014 ; 164 : 962 – 5 . Google Scholar Crossref Search ADS PubMed von Rhein M , Buchmann A , Hagmann C , Huber R , Klaver P , Knirsch W et al. . Brain volumes predict neurodevelopment in adolescents after surgery for congenital heart disease . Brain 2014 ; 137 : 268 – 76 . Google Scholar Crossref Search ADS PubMed von Rhein M , Buchmann A , Hagmann C , Dave H , Bernet V , Scheer I et al. . Severe Congenital heart defects are associated with global reduction of neonatal brain volumes . J Pediatr 2015 ; 167 : 1259 – 63.e1 . Google Scholar Crossref Search ADS PubMed Welker KM , Patton A . Assessment of normal myelination with magnetic resonance imaging . Semin Neurol 2012 ; 32 : 15 – 28 . Google Scholar Crossref Search ADS PubMed Wickremasinghe AC , Hartman TK , Voigt RG , Katusic SK , Weaver AL , Colby CE et al. . Evaluation of the ability of neurobiological, neurodevelopmental and socio-economic variables to predict cognitive outcome in premature infants . Child Care Health Dev 2012 ; 38 : 683 – 9 . Google Scholar Crossref Search ADS PubMed Winkler AM , Ridgway GR , Webster MA , Smith SM , Nichols TE . Permutation inference for the general linear model . Neuroimage 2014 ; 92 : 381 – 97 . Google Scholar Crossref Search ADS PubMed Witelson SF , Pallie W . Left hemisphere specialization for language in the newborn. Neuroanatomical evidence of asymmetry . Brain 1973 ; 96 : 641 – 6 . Google Scholar Crossref Search ADS PubMed © The Author(s) (2019). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Left temporal plane growth predicts language development in newborns with congenital heart disease
JO - Brain
DO - 10.1093/brain/awz067
DA - 2019-05-01
UR - https://www.deepdyve.com/lp/oxford-university-press/left-temporal-plane-growth-predicts-language-development-in-newborns-f7YX1S3anT
SP - 1270
VL - 142
IS - 5
DP - DeepDyve
ER -