Ultrasound and MRI of fetuses with ventriculomegaly: can cortical development be used to predict postnatal outcome?

Abstract

Objective: The purpose of this study was to assess the conspicuity of brain cortical maturation with sonography and MRI of fetuses referred because of ventriculomegaly and to determine whether sulcal visualization can be used to predict postnatal outcome.

Subjects and methods: Women with 374 fetuses referred because of ventriculomegaly underwent sonography and MRI. Four to six radiologists rated visualization of 19 fissures or sulci. Majority opinion regarding sulcal visualization was compared among fetuses categorized by CNS abnormality: normal, isolated ventriculomegaly, and ventriculomegaly with additional CNS abnormalities. Live-born infants were categorized as having normal or abnormal development. Logistic regression analysis was used to correlate sulcal visualization and postnatal development. A subanalysis was performed with fetuses who had been classified prenatally as having a normal brain or isolated ventriculomegaly.

Results: Cortical sulci were visualized more frequently and at an earlier gestational age with MRI than with ultrasound. In the entire cohort, the odds ratio of normal development ranged from 3.1 to 10.0 whenever the calcarine, parietooccipital, cingulate, superior temporal, precentral, or postcentral sulcus was seen on MR images. In fetuses categorized as having a normal brain or having isolated ventriculomegaly, the odds ratio of normal development ranged from 3.5 to 9.0 whenever the parietooccipital, cingulate, or superior temporal sulcus was seen.

Conclusion: Visualization of the sulci in fetal brains depends on the imaging modality used and the gestational age at imaging. Information regarding sulcal visualization may aid in counseling patients carrying fetuses with ventriculomegaly.

Figures

Fig. 1

Flowchart of patient enrollment, birth outcome, and postnatal outcome.

Fig. 2

Fig. 2a. Sagittal T2 weighted image at 28 weeks gestation showing precentral sulcus (arrowhead) and postcentral sulcus (arrow). Fig. 2b. Sagittal T2 weighted image at 30 weeks gestation showing cingulate sulcus (arrowhead), parieto-occipital sulcus (white arrow) and calcarine sulcus (black arrow).

Fig. 3

Fig. 3a. Sagittal transvaginal US at 30 weeks gestation showing precentral sulcus (arrowhead) and postcentral sulcus (arrow). Fig. 3b. Sagittal tranvaginal US at 29 gestation showing cingulate sulcus (arrowhead), parieto-occipital sulcus (large arrow), and calcarine sulcus (thin arrow).

Fig. 4

Percentage of readings in which each sulcus was visualized on MR and US. More sulci were visualized, and visualized more frequently, on MR than on US.

Fig. 5

Percent agreement on visualization of each sulcus on US and MR based on first 194 fetuses. Percentage of agreement was higher for sulci that were visualized most frequently or least frequently.

Fig. 6. Percentage of subjects with a normal clinical outcome in all patients

Normal developmental outcome was more likely when sulcation patterns were visualized on prenatal MR (right) than in the sample as a whole (18%, dashed line) or when the pattern was not visualized (left). Bars indicate simple percentages with standard error, not adjusted for gestational age at time of imaging. Error bars are 1 standard error. *p

Fig. 7. Percentage of subjects with a normal clinical outcome in patients with a diagnosis of normal or isolated VM

Normal developmental outcome was more likely when sulcation patterns were visualized on prenatal MR (right) than in the sample as a whole (29%, dashed line) or when the pattern was not visualized (left). Bars indicate simple percentages with standard error, not adjusted for gestational age at time of imaging. Error bars are 1 standard error. *p

Fig. 8. Percentage of subjects with a normal clinical outcome in patients with a diagnosis of normal or isolated VM who did not undergo termination, limited to patients imaged at great than 22 weeks GA

Normal developmental outcome was more likely when sulcation patterns were visualized on prenatal MR (right) than in the sample as a whole (37%, dashed line) or when the pattern was not visualized (left). Bars indicate simple percentages with standard error, not adjusted for gestational age at time of imaging. Error bars are 1 standard error. *p