Brainstem–vermis and brainstem–tentorium angles allow accurate categorization of fetal upward rotation of cerebellar vermis

Authors


Abstract

Objective

To evaluate the role of the brainstem–vermis (BV) and brainstem–tentorium (BT) angles in the differential diagnosis of upward rotation of the fetal cerebellar vermis.

Methods

The BV and BT angles were measured retrospectively on median sonographic views of the brain in 31 fetuses at 19–28 weeks' gestation with upward rotation of the cerebellar vermis due to Blake's pouch cyst (n = 12), Dandy–Walker malformation (n = 12) and cerebellar vermian hypoplasia (n = 7). Eighty normal fetuses at 20–24 weeks were included as controls.

Results

In the control group, BV and BT angles were 9.1 ± 3.5° (range, 4–17°) and 29.3 ± 5.8° (range, 21–44°), respectively. The BV angle was significantly increased in each of the three subgroups of anomalies: Blake's pouch cyst (23 ± 2.8°; range, 19–26°), vermian hypoplasia (34.9 ± 5.4°; range, 24–40°) and Dandy–Walker malformation (63.5 ± 17.6°; range, 45–112°), the angle increasing with increasing severity of the condition. The BT angle had a similar pattern but there was overlap among the different groups.

Conclusion

The BV angle and, to a lesser degree, the BT angle are simple and reproducible measurements that provide valuable additional information for the categorization of upward rotation of the fetal cerebellar vermis. From mid gestation, a BV angle > 45° is strongly suggestive of a Dandy–Walker malformation, while a measurement < 30° favors the diagnosis of a Blake's pouch cyst. Copyright © 2012 ISUOG. Published by John Wiley & Sons, Ltd.

Introduction

Fetal posterior fossa fluid collections associated with upward rotation of the cerebellar vermis range from benign asymptomatic conditions to severe abnormalities associated with neurological impairment1–9. The most frequent of these anomalies, Blake's pouch cyst, vermian hypoplasia and Dandy–Walker malformation, have a similar sonographic appearance but a very different prognosis4. A specific diagnosis with either ultrasound or magnetic resonance imaging is possible, but is frequently difficult and relies mostly upon subjective criteria1, 4–9.

We aimed to evaluate the diagnostic contribution of an objective approach based on the brainstem–vermis (BV) angle and the brainstem–tentorium (BT) angle, which we have demonstrated previously to be feasible and reproducible sonographic measurements10.

Methods

BV and BT angles were measured retrospectively in two referral centers for prenatal diagnosis, in fetuses with posterior fossa fluid collections associated with upward rotation of the cerebellar vermis (study group) retrieved from a series previously reported4 and prospectively in consecutive normal fetuses at mid gestation (controls). Measurements were obtained from median views of the fetal brain as previously described (Figure 1)10. Inclusion criteria for the study group were availability of digital images and/or ultrasound volumes of good quality as well as detailed postnatal follow-up. Three-dimensional (3D) ultrasound volumes and two-dimensional (2D) digital images demonstrating a median view of the brain were used for measurements in the study group, and 3D volumes were used for controls, utilizing 4D View 9.0 (GE Healthcare, Milan, Italy) and Adobe Photoshop 6.0 (Adobe Systems Incorporated, San Jose, CA, USA) software, respectively. All measurements were performed by two operators (E.C., P.V.). Statistical analysis was performed by calculating means and SDs. Groups were compared using the Mann–Whitney U-test.

Figure 1.

Measurement of brainstem–vermis (BV) and brainstem–tentorium (BT) angles. (a) A median view of the fetal brain is obtained (in this case after acquisition of an ultrasound volume starting from an axial view) and the main anatomic landmarks are identified. (b) A line is drawn tangentially to the dorsal aspect of the brain stem and a second line is drawn tangentially to the ventral contour of the cerebellar vermis; the interposed angle (1) is the BV angle; the BT angle (2) is measured between the first line and a third line tangential to the tentorium.

Results

The study group included 31 fetuses at 19–28 weeks' gestation with posterior fossa fluid collections (12 with Blake's pouch cyst, 12 with Dandy–Walker malformation and seven with vermian hypoplasia) (Figure 2). The control group comprised 80 normal fetuses at 20–24 weeks' gestation. BV and BT angle measurements are reported in Table 1. Controls always had a BV angle < 18° and a BT angle < 45°. The BV angle was significantly increased in each of the three subgroups of anomalies (Figure 3, Table 2), the angle increasing with increasing severity of the condition. The BT angle demonstrated a similar pattern, but there was more overlapping among groups (Figure 4, Table 2).

Figure 2.

Measurement of brainstem–vermis (BV) angle (1) and brainstem–tentorium (BT) angle (2) in fetuses with: (a) Blake's pouch cyst; (b) cerebellar vermis hypoplasia; and (c) Dandy–Walker malformation. The BV angles are 26°, 39° and 73°, respectively and the BT angles are 45°, 50° and 66°. These images were obtained from three-dimensional ultrasound volumes acquired originally by positioning the probe along the posterior fontanelle. Corresponding images from the same cases obtained by acquiring the volumes with a transabdominal axial approach are also provided (d,e,f), to demonstrate the excellent correlation between the two approaches.

Figure 3.

Box-and-whisker plot of distribution of brainstem–vermis angle in controls and in fetuses with upward rotation of the cerebellar vermis. Medians are indicated by a line inside each box, 25th and 75th percentiles by box limits and 5th and 95th percentiles by lower and upper bars, respectively.

Figure 4.

Box-and-whisker plot of distribution of brainstem–tentorium angle in controls and in fetuses with upward rotation of the cerebellar vermis. Medians are indicated by a line inside each box, 25th and 75th percentiles by box limits and 5th and 95th percentiles by lower and upper bars, respectively.

Table 1. Brainstem–vermis (BV) and brainstem–tentorium (BT) angles in fetuses with upward rotation of the cerebellar vermis and in controls
 BV angle (°)BT angle (°)
Ultrasound findingsnMeanSDRangeMeanSDRange
Controls809.13.54–1729.35.821–44
Blake's pouch cyst1223.02.819–2642.27.132–52
Vermian hypoplasia734.95.424–4052.17.045–66
Dandy–Walker malformation1263.517.645–11267.215.151–112
Table 2. Statistical comparison of brainstem–vermis (BV) and brainstem–tentorium (BT) angles in controls and in fetuses with upward rotation of the cerebellar vermis
 P (Mann–Whitney U-test)
Comparison*BV angleBT angle
  • *

    Group with smaller angle vs. group with larger angle.

Controls vs Blake's pouch cyst fetuses< 0.00000005< 0.000005
Controls vs Dandy–Walker fetuses< 0.00000005< 0.00000005
Controls vs vermian hypoplasia fetuses< 0.00005< 0.00005
Blake's pouch cyst vs Dandy–Walker fetuses< 0.00005< 0.00005
Blake's pouch cyst vs vermian hypoplasia fetuses< 0.0050.01
Vermian hypoplasia vs Dandy–Walker fetuses< 0.0005< 0.005

Discussion

Our results suggest that measurement of the BV angle discriminates accurately posterior fossa fluid collections associated with upward rotation of the cerebellum.

In the late first trimester, the fourth ventricle is large and a relatively small cerebellum is located on top of it. In the following weeks, the cerebellum grows to enfold completely the fourth ventricle. However, a small finger-like appendage of the fourth ventricle, the Blake's pouch, is frequently seen protruding into the cisterna magna, caudal to the cerebellum8. It has been suggested that there is a continuum of anatomic anomalies involving the fourth ventricle–Blake's pouch complex8. The one with least clinical impact among these anomalies is the Blake's pouch cyst, a persistence of the Blake's pouch that results in an isolated superior displacement of the cerebellar vermis. At the other end of the spectrum lies the Dandy–Walker malformation, in which the upward displacement of a normal to hypoplastic vermis is associated with enlargement of the cisterna magna. In vermian hypoplasia (previously referred to as Dandy–Walker variant) the cisterna magna is of normal size and the vermis is small and frequently (although not always) rotated upward. Distinguishing these three entities is important, because of their different prognoses: a Blake's pouch cyst is a risk factor for anatomic and chromosomal anomalies but when isolated is probably a normal variant, while Dandy–Walker malformation and vermian hypoplasia are true malformations frequently associated with abnormal neurodevelopment4. The differential diagnosis depends upon visualization of the tentorium and/or torcular (normal position in Blake's pouch cyst and vermian hypoplasia, upward displacement in Dandy–Walker malformation) and appearance of the vermis (intact in Blake's pouch cyst, hypoplastic in vermian hypoplasia, normal to hypoplastic in Dandy–Walker malformation)1, 4–7. These findings can be demonstrated in utero with sonography and/or magnetic resonance imaging, but they are subjective and even in expert hands may be difficult to interpret, particularly early in gestation4.

In our series, the BV angle discriminated accurately this group of anomalies. In normal fetuses at mid-gestation the angle was always < 18°. In fetuses with Blake's pouch cyst it was always < 30°, in those with Dandy–Walker malformation it was consistently > 45°, while vermian hypoplasia had intermediate values (Figure 3).

The BT angle was increased in the study group compared with controls, but there was overlapping among the different types of anomalies, limiting its diagnostic value (Figure 4).

We acknowledge the limitations of our study. The number of abnormal cases was relatively small and they were investigated retrospectively. Further experience is certainly needed. Nevertheless, the spread of measurements between normal and abnormal cases and among the different categories of abnormalities suggest that the BV angle may prove important in the differential diagnosis of fetal posterior fossa fluid collections, at least when used in combination with the traditional criteria1, 4, 5, 7. After 20 weeks' gestation, a measurement > 45° is strongly indicative of a Dandy–Walker malformation, while a value < 30° favors the diagnosis of Blake's pouch cyst. An intermediate value in our series was associated with vermian hypoplasia, but the number of cases was limited and caution is warranted in making this difficult diagnosis4, 6.

In conclusion, we suggest that the BV angle and, to a lesser extent, the BT angle are objective findings useful in differentiating fetal posterior fossa fluid collections, that are sonographically similar but carry a very different prognosis.

Ancillary