Association of fetal cerebral growth and sulcal development with 2‐year neurodevelopmental outcome in children with prenatally diagnosed mild‐to‐moderate isolated ventriculomegaly

Isolated cerebral ventriculomegaly (IVM) is the most common fetal brain anomaly, affecting 0.3–1.5 per 1000 births1. It is defined by an atrial diameter of the lateral ventricles > 10 mm, with severity classified as mild (≥ 10 mm and < 13 mm), moderate (≥ 13 mm and < 15 mm) or severe (≥ 15 mm). Neurodevelopmental outcomes in IVM range from age-appropriate to significant delay, with the incidence of delay reported to be 5–25%. However, prediction of outcomes in IVM remains challenging, and no quantitative tools are currently in clinical use. While larger ventricular size has been linked to poorer neurodevelopmental outcomes in cases of severe IVM1, studies of mild-to-moderate IVM cases show no consistent association between ventricular diameter or volume and neurodevelopmental outcomes2, 3. Recent studies highlight cortical folding and regionalization as critical prenatal processes for brain organization and function, and therefore neurodevelopment4, 5. In our prior work, fetuses with IVM demonstrated increased brain and cortical volumes, along with disrupted sulcal position patterns, compared with controls6. Since atypical sulcal position patterns are linked to poorer neurodevelopmental outcomes in various neurological disorders7, we hypothesized that altered fetal regional brain growth and abnormal sulcal position patterns in IVM are associated with neurodevelopmental delay at around 2 years of age. To test this, we prospectively studied a cohort of fetuses diagnosed prenatally with IVM and reported the postnatal outcomes. Building on our prospective fetus-to-infant cohort study published previously6, we extended this work to include more cases of IVM with neurodevelopmental outcome data. Brain magnetic resonance imaging (MRI) data of fetuses with mild-to-moderate IVM (10 mm ≤ atrial diameter < 15 mm) were processed and regional volumetric and sulcal position pattern analyses were conducted according to the methodology employed in our previous study6 (Appendix S1). Regional volumes of the left and right cortical plate, subcortical parenchyma, cerebellar hemispheres, cerebellar vermis, brainstem and lateral, third and fourth ventricles were computed based on automated and subsequent manual segmentation. For both individual cases of IVM and normal template brains, sulcal basins and pits (the deepest part of the sulcal basin) were identified automatically based on the curvature map, and three-dimensional (3D) sulcal positions and intersulcal positional relationships were determined in the left and right whole hemispheres (Figure 1). Sulcal position pattern analysis computed the similarity index of the sulcal position pattern between individual fetal cases and matched normal template brains. The similarity index ranges from 0 to 1, with a lower value indicating a greater deviation from the normal sulcal position pattern6. Neurodevelopmental outcome was assessed through a review of medical records, using the most recent available clinical documentation by pediatric neurologists and pediatricians, and recording the presence or absence of a documented diagnosis of motor or cognitive developmental delay or behavioral problems. To compare fetal MRI measurements between age-appropriate and delayed neurodevelopment groups, we conducted statistical analyses by utilizing a non-linear regression model comparison (for regional volume) and a non-parametric Mann–Whitney U-test (for sulcal similarity index). P< 0.05 was considered statistically significant. Twenty-six pregnant women were recruited at Tufts Medical Center, Boston, MA, USA, and Jikei University Hospital, Tokyo, Japan (Table S1). Fetal MRI was obtained at a mean ± SD gestational age of 27.6 ± 5.6 weeks and mean ± SD maternal age at MRI was 28.4 ± 5.9 years. MRI protocols at the two institutions are listed in Table S2. Median gestational age at birth was 39.2 (range, 37.1–42.1) weeks. Neonatal MRI of all cases confirmed the diagnosis of IVM. At follow-up (mean infant age, 24.6 ± 12.5 months), 17/26 children had age-appropriate neurodevelopment and 9/26 had neurodevelopmental delay (five with speech delay, two with motor delay, one with behavioral problems and one with speech delay and behavioral problems). Outcome assessments included the Bayley Scales of Infant and Toddler Development, Third Edition (n = 8), Ages & Stages Questionnaires (n = 3), pediatrician evaluation (n = 5) and pediatric neurologist evaluation (n = 10). On regional volumetric analysis, children with delayed neurodevelopment had significantly larger bilateral cortical plate volumes (left, P = 0.020; right, P = 0.012) and right cerebellar hemisphere volume (P = 0.017) compared with those with age-appropriate neurodevelopment (Figure 2a,b,f). No significant differences between the groups were found in the subcortical parenchyma, left cerebellum or lateral ventricles (P > 0.15) (Figure 2c–e,g,h). All 26 cases underwent sulcal position pattern analysis, however, 15 were excluded due to specific criteria: gestational age < 24 weeks and therefore minimal sulcation (n = 8), gestational age > 35 weeks and therefore overly complex sulcation (n = 2) or poor 3D surface reconstruction (n = 5) (Figure S1). The final analysis included seven children with age-appropriate neurodevelopment and four with delayed neurodevelopment. Delayed cases showed significantly lower similarity indices in left-hemisphere sulcal position patterns: whole pattern, corresponding sulcal basins, and intersulcal relationship (all P = 0.042) (Figure 3a–c, Table S3). No significant differences were observed in the sulcal position patterns of the right hemisphere (all P = 0.927) (Figure 3d–f, Table S3). These results were not affected by site effects, as shown by Figure S2. This work found that, in fetuses with IVM, accelerated bilateral cortical plate and right cerebellar growth, along with abnormal left sulcal position pattern during the second and third trimesters, were significantly associated with neurodevelopmental delay at approximately 2 years of age. These findings suggest quantitative fetal neuroimaging may help predict neurodevelopmental outcomes. Prior studies of IVM have focused on ventricular diameter, ventricular volume and intracranial size, with inconsistent correlation with neurodevelopment2, 3. Our regional volumetric analysis showed that cortical and cerebellar growth was accelerated in the neurodevelopmental delay group, suggesting more pronounced pathology and possible alterations in neurogenesis. This is also consistent with a recent study that identified potential correlations between ventricular volume and autistic traits at 2 years of age4. Abnormal sulcal position pattern also correlated with neurodevelopmental delay, highlighting potential disruption in cortical regionalization. Similar alterations have been observed in agenesis of the corpus callosum7, Down syndrome and other developmental disorders, and may persist into childhood. These results underscore the clinical importance of sulcal features, which are often missed on conventional MRI scans, and support the use of region-specific sulcal measures in future multicenter studies. Limitations include variability in manual segmentation despite generally high inter-rater Dice coefficients in our prior work. Sulcal position pattern analyses were also limited to scans conducted after 23 weeks' and before 35 weeks' gestation, reducing sample size, and additional exclusions occurred due to poor 3D reconstructions. Furthermore, normative modeling using a large normative dataset may enhance the sensitivity for detecting aberrant brain development in IVM. Neonatal MRI data are yet to be analyzed in a future study to provide longitudinal insight into regional growth and sulcation. Outcome data relied on chart review, with variable timing and potential bias toward children with neurodevelopmental concerns. The higher-than-expected rate of neurodevelopmental delay in all cases with available outcome data may reflect this selection. Implementing standardized neurodevelopmental testing is therefore essential. Outcomes were also only assessed at 2 years of age, leaving uncertainty about persistence into later life. Larger cohorts with standardized, long-term follow-up to school age are needed. H. Yun, Fetal-Neonatal Neuroimaging and Developmental Science Center, Division of Newborn Medicine Boston Children's Hospital, and Department of Pediatrics, Harvard Medical School, Boston, MA, USA G. Graham, Obstetrics and Gynecology, Tufts Medical Center, Boston, MA, USA R. Kitano, Mother Infant Research Institute, Tufts Medical Center, Boston, MA, USA S. Akiyama, Mother Infant Research Institute, Tufts Medical Center, Boston, MA, USA The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Appendix S1 Supplementary methods. Table S1 Subject demographics of 26 cases of isolated cerebral ventriculomegaly included in study. Table S2 Fetal magnetic resonance imaging parameters from the two study institutions. Table S3 Sulcal position pattern similarity indices of age-appropriate vs delayed neurodevelopment groups compared with matched normal template brains. Figure S1 Flowchart showing case selection and exclusion through imaging analysis process. Figure S2 Distribution of site-specific study data in regional volumetric analysis of left cortical plate (a) and sulcal position pattern similarity analysis compared with matched normal template brains of left whole sulcal pattern (b) in fetuses with isolated ventriculomegaly who presented with delayed vs age-appropriate neurodevelopment at 2 years of age. Data points derived from Jikei University (JU) cases (JU11 and JU13) are labelled with arrows. The remaining data points were derived from cases at Tufts Medical Center. Distribution of JU datapoints appeared not to cause site effects, as they are evenly distributed between the delayed and age-appropriate neurodevelopment groups. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.