Prenatal Diagnosis of Fetal Aqueductal Stenosis: A Multicenter Prospective Observational Study through the North American Fetal Therapy Network

INTRODUCTION

A critical component of an evidence-based reassessment of in-utero intervention for fetal aqueductal stenosis (fetal AS) is determining if the prenatal diagnosis can be accurately made at a gestational age amenable to in-utero intervention.

METHODS

A multicenter, prospective, observational study was conducted through the North American Fetal Therapy Network (NAFTNet). Pregnancies complicated by severe central nervous system (CNS) ventriculomegaly (lateral ventricle diameter >15 mm) not secondary to a primary diagnosis (myelomeningocele, encephalocele, etc.) were recruited at diagnosis. Imaging and laboratory findings were recorded in an online REDCap database. After evaluation, investigators were asked to render their degree of confidence in the diagnosis of fetal AS. The prenatal diagnosis was compared to the postnatal diagnosis obtained through neonatal neuroimaging. Performance characteristics of ultrasound and magnetic resonance imaging (MRI) were calculated, as was the mean gestational age at diagnosis.

RESULTS

Between April 2015 and October 2022, eleven NAFTNet centers contributed 64 subjects with severe fetal CNS ventriculomegaly. Of these, 56 had both prenatal and postnatal diagnoses recorded. Ultrasound revealed 32 fetal AS true positives, 4 false positives, 7 false negatives, and 13 true negatives, rendering a sensitivity of 0.82, a specificity of 0.76, a positive predictive value of 0.89, and a negative predictive value of 0.65. The mean gestational age at diagnosis by ultrasound was 25.5 weeks (std +/- 4.7 weeks). The proportion of agreement (true positive + true negative/n) was highest at 24 weeks gestation. For fetal MRI (n = 35), the sensitivity for fetal AS was 0.95, specificity was 0.69, positive predictive value was 0.84, and negative predictive value was 0.90. MRI was performed at 25 weeks on average.

CONCLUSION

The prenatal diagnosis of fetal AS can be made with accuracy at a gestational age potentially amenable to in-utero intervention. Only 7% of subjects were incorrectly diagnosed prenatally with fetal AS by ultrasound and 11% by MRI. Diagnostic accuracy of fetal AS will likely improve with increased experience.

Comparison of fetal and neonatal sonographic measurements of ventricular size in second- and third-trimester fetuses with or without ventriculomegaly: cross-sectional three-dimensional ultrasound study

OBJECTIVES

To assess, in a population comprising normal fetuses and fetuses with primary or post-hemorrhagic ventriculomegaly, the reproducibility of measurement of neonatal ultrasound indices in the fetus and to compare the performance of various cut-offs of these parameters to diagnose ventriculomegaly and classify its severity.

METHODS

This was a retrospective cross-sectional study including 182 singleton fetuses assessed by transvaginal neurosonography. The sample populations included 116 normal fetuses and 66 fetuses with primary (n = 56) or post-hemorrhagic (n = 10) ventriculomegaly. In all cases, the atrial width (AW) was measured according to standard protocols and the findings were compared with four sonographic indices developed in the neonate: the anterior horn width (AHW), the ventricular index (VI), the thalamo-occipital distance (TOD) and the fronto-occipital horn ratio (FOHR). Reproducibility of measurements was assessed using the intraclass correlation coefficient (ICC) and diagnostic accuracy of the neonatal indices was assessed against AW using areas under the receiver-operating-characteristics curves (AUC).

RESULTS

The intra- and interoperator reproducibility of measurement of AW and the neonatal measurements was excellent, with ICCs > 0.99 for all measures. The association in the fetus of all four variables developed in the neonate with the degree of ventriculomegaly as defined by the AW was strong for severe ventriculomegaly (AW > 15.0 mm; all AUC > 0.95), whereas the separation of cases with mild ventriculomegaly (AW, 10.0-15.0 mm) from those with normal AW (< 10.0 mm) was less effective.

CONCLUSIONS

When applied in the fetus, all four indices of ventriculomegaly developed in neonates (AHW, VI, TOD, FOHR) were associated strongly with fetal AW when the AW measurement indicated severe fetal ventriculomegaly. However, for mild ventriculomegaly, the association was weaker, probably due to the fact that, in the fetus, mild ventriculomegaly is not caused by obstruction of the ventricular system. Considering the similar performance of the four neonatal variables and the technical issues involved in determination of TOD and FOHR in the fetus, use of VI and AHW is preferred. © 2022 International Society of Ultrasound in Obstetrics and Gynecology.

Fetal aqueductal stenosis: Prenatal diagnosis and intervention

Fetal severe central nervous system ventriculomegaly is associated with poor neurologic outcomes, usually driven by a primary malformation, deformation, or disruption of brain parenchyma. In utero shunting of excess cerebrospinal fluid (CSF) in hopes of improving neurologic outcomes was attempted in the 1980s but was abandoned due to perceived lack of effect, likely due to technological limitations of the time that precluded proper patient selection. Little progress on the antenatal management of severe ventriculomegaly has been made in the intervening decades. A multidisciplinary, evidence-based reassessment of ventriculoamniotic shunting for isolated fetal aqueductal stenosis (FAS), a unique form of severe ventriculomegaly (supratentorial intracranial hypertension), is currently underway. An accurate diagnosis of FAS must precede in utero intervention. Magnetic resonance imaging (MRI) will be an excellent adjunct to high-resolution prenatal ultrasound and next-generation genetic testing to correctly diagnose FAS in a timely fashion while excluding other intracranial and extracranial anomalies. This manuscript will briefly discuss the history, current management, and future directions of the prenatal diagnosis and potential intervention for FAS.

The many faces of prenatal imaging diagnosis of primitive aqueduct obstruction

OBJECTIVE

To describe the different prenatal imaging patterns related to primary aqueduct obstruction throughout pregnancy and their impact on prenatal counseling.

METHOD

Retrospective review of consecutive prenatal cases of isolated aqueduct obstruction diagnosed over a seven-year period (2010-2016). Prenatal imaging findings, postnatal imaging, pathological data and postnatal outcome, were analyzed.

RESULTS

Twelve cases were included. In four cases, termination of pregnancy was performed, due to early severe ventriculomegaly in three cases suggestive of complete aqueduct obstruction. In eight cases in which pregnancy was continued, three different evolving imaging patterns were demonstrated. In three cases the ventriculomegaly evolved rapidly during third trimester and were subsequently associated with abnormal white matter changes on fetal MRI. Theses cases led to premature delivery in two cases and early surgical care in all, with focal cystic parenchymal damage shown on follow-up MRI in two cases. Slowly evolving ventriculomegaly in three cases diagnosed in the second and the third trimester which required delayed surgery during the first year of life. Stable ventriculomegaly in two cases which did not require any surgical procedure.

CONCLUSION

The diagnosis of primary aqueduct obstruction may be based on different prenatal imaging patterns that include either severe early ventriculomegaly, stable, slowly or rapidly evolving ventriculomegaly.

The third ventricle of the human fetal brain: Normative data and pathologic correlation. A 3D transvaginal neurosonography study

OBJECTIVE

The objective of the study are to describe (a) the technical aspects and (b) the anatomical boundaries of the fetal third ventricle (3V) on the midsagittal sonographic view and to assess (c) different biometric parameters in normal and abnormal fetuses and (d) and their reproducibility.

METHODS

This study included 67 normal and 50 CNS anomalies fetuses which include (1) obstructive severe ventriculomegaly (SVM; atrial width ≥ 15 mm), (2) moderate ventriculomegaly (10-14.9 mm), and (3) corpus callosum agenesis (ACC). All underwent transvaginal 3D neurosonography of the midsagittal view of the 3V. The following parameters were measured: area, perimeter, craniocaudal and anteroposterior (AP) diameters, interthalamic adhesion diameter (ITAD), wedge angle, and the ratio between the last 2 variables (ITAD/WA). Repeatability was also assessed.

RESULTS

The ITAD and the ITAD/WA are significantly different between normal fetuses and the SVM (P ≤ .001). Interthalamic adhesion diameter of ≤7.1 mm is able to identify SVM with 98.6% accuracy (CI: 0.92-0.99). In ACC cases, the AP diameter is significantly shorter than both normal fetuses and ventriculomegaly. Intraobserver/interobserver reliability was good for most variables.

CONCLUSIONS

Transvaginal neurosonography enables visualization of the normal and abnormal fetal third ventricle. An ITAD <7.1 identifies aqueductal stenosis as the likely etiology of severe ventriculomegaly with an accuracy of 98.6%.

The suprapineal recess of the third ventricle: an anatomic study with magnetic resonance imaging

PURPOSE

The suprapineal recess (SPR) is a small, backward extension of the third ventricle. Few radiological studies have investigated the morphology of the SPR. Here, we explore the SPR with magnetic resonance (MR) imaging.

METHODS

A total of 124 patients underwent thin-slice MR imaging examinations with T2-weighted imaging and the constructive interference steady-state (CISS) sequence. Imaging data were transferred to a workstation for analysis.

RESULTS

The pineal gland (P) was delineated in 99% of the patients on T2-weighted imaging and 100% of the patients on the CISS sequence. In contrast, the SPR was identified in 27% of the patients on T2-weighted imaging and 82% of the patients on the CISS sequence. The location of the P relative to the lowest point of the splenium was roughly classified into two types. Of them, the anterior P location was the more frequent type and observed in 73% of the patients. The angle formed by the roof and floor of the SPR showed remarkable interindividual diversity. A membranous posterior extension with variable length, spanning between the posterosuperior margin of the P and Galenic complex was found in 55% of the identified SPRs on T2-weighted imaging and 45% on the CISS sequence.

CONCLUSIONS

The SPR is a distinct structure with diversity in appearance among individuals but commonly extends posterior to the P. High-resolution MR imaging is useful for delineating the SPR in vivo.