The Dandy-Walker malformation complex refers to a group of congenital central nervous system malformations that primarily involve the cerebellum and adjacent structures. The diagnostic nomenclature in this field is confusing, in part because this complex may be regarded as a spectrum of posterior fossa abnormalities, which includes dilatation of the fourth ventricle, hypoplasia to aplasia of the cerebellar vermis, rotation of the cerebellar vermis, and elevation of the tentorium and transverse sinus.1–4 Dandy-Walker malformation complex is associated with an increased incidence of developmental disability, but it has been suggested that long-term prognosis correlates best with other anomalies.5–14 Therefore, diagnosis of Dandy-Walker complex necessitates a diligent search for additional anomalies in the central nervous system (CNS, eg, ventriculomegaly) and other organs. Cytogenetic evaluation is also indicated because Dandy-Walker malformation complex can occur in the context of various chromosomal disorders.8,12,15,16 Because of concern about neurologic dysfunction and disability, many parents terminate pregnancies electively after an ultrasound diagnosis of Dandy-Walker malformation complex, even without other sonographic or cytogenetic abnormalities. Hence, accurate prenatal diagnosis of this condition is imperative.
Criteria used to diagnose Dandy-Walker complex by prenatal ultrasound examination include enlargement of the cisterna magna (anterior-posterior depth ≥ 10 mm) and aplasia/hypoplasia of the cerebellar vermis, as evidenced by a sonolucent gap between the cerebellar hemispheres in the transcerebellar view.6,7,13,17,18 This gap is contiguous with the enlarged fourth ventricle and is sometimes referred to as “continuity” or “communication” between the cisterna magna and fourth ventricle.9,12,19,20
Anatomic pathologic examination remains the gold standard for establishing the validity of many diagnostic test results. In general, reported concordance between CNS malformations diagnosed prenatally by ultrasound examination and neuropathologic findings at autopsy is high.21–23 Unfortunately, this is not the case for Dandy-Walker malformation complex, and the rate of confirmation of prenatal ultrasound diagnoses of this complex by either neuropathology or postmortem radiology is highly variable (Table 1).12,21–26 Failure to confirm Dandy-Walker complex at autopsy is particularly disconcerting when the sonographic diagnosis provided the sole basis for elective termination.
In this study, we assessed concordance between the anatomical pathology at autopsy and the prenatal sonographic observations for a series of 44 fetuses diagnosed with Dandy-Walker malformation complex by prenatal ultrasound examination. In addition to comparing the original ultrasound and pathology results, we performed a “blinded” review of ultrasound and neuropathology images to seek explanations for diagnostic discordance.
MATERIALS AND METHODS
A total of 1,457 consecutive fetal and pediatric autopsy records from Children's Hospital and Regional Medical Center and University of Washington Medical Center between 1995 and 2003 were reviewed to identify 44 with a prenatal ultrasound diagnosis of Dandy-Walker malformation complex. Initial ultrasound examinations were performed in multiple facilities, and 37 of 44 cases had confirmatory sonographic examinations at either University of Washington Medical Center or Swedish Hospital and Medical Center. Diagnostic prenatal sonographic images of the posterior fossa were available from 31 of these cases (mean gestational age 22.0 ± 6.0 weeks). In the remaining 13, either ultrasound images of the posterior fossa could not be obtained or the available images were inadequate for the diagnosis of Dandy-Walker malformation complex. Autopsy reports, including gross and microscopic descriptions, were examined from all 44 cases. In 34 cases, at least one postmortem photograph and/or histological slide of the cerebellum was available for review (mean gestational age 24.1 ± 7.1 weeks). Age-matched (± 1 week) controls were randomly selected on a one-to-one basis from autopsy cases, with no evidence of central nervous system abnormalities detected by prenatal ultrasonography or autopsy (31 control ultrasound cases, mean gestational age 22.7 ± 5.6 weeks; 36 control pathology cases, mean gestational age 24.0 ± 6.0 weeks). The prenatal sonograms and histologic slides of the cerebellum from the control cases were intermixed with the study cases and reviewed in a blinded manner to reduce diagnostic bias. Institutional review boards at all 3 academic institutions approved this study.
Two radiologists independently evaluated the ultrasound images from each case and controls, blinded to the original sonographic and autopsy interpretations. The sonograms were all acquired via transabdominal sonography, and each set included slightly angled axial views through the thalami, cerebral peduncles, and posterior fossa (transcerebellar view) to assess the depth of the cisterna magna. Sonographic diagnosis of Dandy-Walker malformation complex from the standard transcerebellar images required enlargement of the cisterna magna (≥ 10 mm) with an abnormal gap between the cerebellar hemispheres, which was interpreted as hypoplasia, aplasia, or anterior rotation of the cerebellar vermis.6,7,13,17,18 If the gap appeared to occupy the entire rostral-caudal length of the posterior fossa, vermian aplasia was inferred. In cases that included an axial view through the lateral ventricles (transventricular view), ventricular dilatation was defined as a transverse atrial diameter greater than 10 mm.27 Additional sonographic features that were assessed when present included the transverse cerebellar diameter, cerebellar height, fourth ventricle shape and size, height and width of the vermis, and splaying of the cerebellar hemispheres. After blinded review, cases were designated as “Dandy-Walker malformation complex,” “not Dandy-Walker malformation complex,” or “indeterminate”.
Two fetal pathologists reviewed the gross photographs and/or histological slides of the posterior fossa and cerebellum, blinded to the ultrasound findings and original pathologic diagnoses. Each case was designated as “Dandy-Walker malformation complex,” “not Dandy-Walker malformation complex,” or “indeterminate” (insufficient archival images available to establish an unequivocal diagnosis). Neuropathologic diagnosis of Dandy-Walker complex required definite evidence of aplasia or hypoplasia of the cerebellar vermis in a gross photograph and/or histological section. Additional features sometimes noted in the pathology reports and occasionally preserved in postmortem photographs included a posterior fossa “cyst” or enlargement of the fourth ventricle, elevation of the tentorium, and anterior rotation of the vermis. Anterior rotation of the vermis without hypoplasia or aplasia of the vermis was not considered diagnostic of Dandy-Walker malformation complex and was not observed in this series. The Fisher exact test or the Mann-Whitney test was used to compare data obtained from concordant and discordant cases, and 95% confidence intervals were used to quantify uncertainty.
To evaluate the concordance between prenatal ultrasound diagnosis of Dandy-Walker malformation complex and autopsy pathology, we reviewed a series of autopsy cases. A total of 44 cases with a prenatal ultrasound diagnosis of Dandy-Walker malformation complex were identified between 1995 and 2003. The clinical, sonographic, and pathological findings are listed in Table 2. Nine cases had multiple ultrasound examinations, and a single patient had a prenatal magnetic resonance imaging (MRI) examination. At autopsy, 86% of the cases were fetuses between the ages of 16 and 32 weeks. The remaining cases included 6 liveborn infants who died within six weeks of birth. Abnormalities outside the CNS were identified by ultrasound examination in 30 of 44 cases (68%) and included congenital diaphragmatic hernia (4), renal anomalies (4), congenital heart defects (7), extremity anomalies such as clubfoot (2), and hydrops (3). Additional CNS abnormalities identified by ultrasonography were present in 16 of 44 cases (36%), including ventriculomegaly (12) and agenesis of the corpus callosum (1). Chromosomal analysis was performed in 40 cases, and a total of 13 cases were identified with abnormal karyotypes, including trisomy 13 (2), partial trisomy 13 (1), trisomy 21 (1), mosaic monosomy X (2), and triploidy (1). One fetus had a known exposure to isotretinoin in utero, a putative teratogenic cause of Dandy-Walker malformation complex.28 Excluding intrauterine growth restriction (1), mild ventriculomegaly (1), or clubfoot (1), Dandy-Walker malformation complex was the only major prenatal sonographic finding for 9 cases and the sole medical reason for elective termination in 6 cases.
The 44 cases were divided into “concordant,” “discordant,” and “indeterminate” based on their original autopsy results (Table 2). The original autopsy confirmed Dandy-Walker malformation complex in 18 cases (concordant, 41%), but found no pathologic evidence of this complex in 17 cases (discordant, 39%). Nine cases (20%) were considered indeterminate. The latter could not be evaluated adequately at autopsy secondary to autolysis or other postmortem artifacts. Based on the combination of discordant and indeterminate cases, 59% (26/44, 95% confidence interval [CI] 44–72) of the original ultrasound diagnoses were not confirmed at autopsy. No significant differences in age or frequency of abnormal karyotype or other anomalies were observed between the concordant and discordant groups. Among the 6 cases with isolated Dandy-Walker complex, the autopsy results were concordant for 2, discordant for 2, and indeterminate for 2. The concordance rates were similar for cases with a documented prenatal ultrasound examination at University of Washington Medical Center or Swedish Hospital and Medical Center (41%, 95% CI 26–56) and an outside hospital (43%, 95% CI 16–75).
The method of termination was an important factor affecting the ability of the pathologist to assess the anatomical features of the posterior fossa and its contents. Termination of pregnancy was performed in 32 cases: by induction of labor in 21 cases, by induction after potassium chloride (KCl) injection in 9 cases, and by dilation and evacuation in 2 cases. The latter 2 cases were too disrupted to evaluate brain anatomy. The use of KCl was associated with an indeterminate pathological diagnosis. Nearly half (45%, 95% CI 19–74) of cases delivered after KCl administration were indeterminate because of autolysis, as opposed to 5% (95% CI 0.8–23) of cases terminated by induction of labor without prior KCl injection (P < .019, Fisher exact test).
A blinded review of sonographic and pathologic images was conducted to better understand the possible reasons for the poor concordance between the prenatal ultrasound diagnoses of Dandy-Walker malformation complex and the autopsy results. Review of ultrasound images from 31 cases with a prenatal diagnosis of Dandy-Walker complex confirmed the diagnosis in 27 cases and was indeterminate for the remaining 4 cases. The indeterminate cases included 2 deemed too immature to confidently evaluate (gestational age < 16 weeks), one with global cerebellar hypoplasia, and one with diagnostic disagreement between the 2 reviewers. Based on the original autopsy diagnosis, the cases with a confirmed ultrasonographic diagnosis of Dandy-Walker malformation complex were divided into concordant and discordant (Table 3). Five cases had an original autopsy diagnosis of indeterminate, and these cases were excluded. Severity of the vermian defect, as assessed sonographically, correlated with pathologic confirmation of Dandy-Walker malformation complex (P < .001 Fisher exact test). Although not statistically significant, the mean cisterna magna depth tended to be greater for the concordant than the discordant cases, (14.4 ± 12.4 mm versus 8.5 ± 5.1 mm, respectively), and the depth of the cisterna magna was 10 mm or greater in 58% (95% CI 32–81) of concordant cases in contrast to 30% (95% CI 11–60) of discordant cases (P = .185, Fisher exact test). No false-positive diagnoses of Dandy-Walker malformation complex were made after blinded review of 31 age-matched control ultrasound films that were intermixed with the Dandy-Walker cases.
All available pathology data from cases with a prenatal ultrasound diagnosis of Dandy-Walker malformation complex and from age-matched control cases were blindly reviewed. Archival autopsy data of sufficient quality to diagnose or rule out Dandy-Walker complex was available in 14 study cases and 4 control cases. Dandy-Walker complex was not diagnosed in any of the control cases. The prenatal sonographic diagnosis of this complex had been confirmed in 11 of the above 14 cases by independent review of their prenatal ultrasound films; ultrasound images were not available for the remaining 3 cases.
These 11 cases (Table 4) represent an important subset of the original series because the archival sonographic and pathology data were well preserved and withstood stringent diagnostic criteria applied by the independent reviewers. Of note, the sonographic-pathologic concordance rate on review, 6 of 11 cases (55%, 95% CI 28–79), was similar to the original concordance rate for the entire series. The ultrasonographic diagnosis of Dandy-Walker malformation complex could not be confirmed by pathology in 5 cases (45%; 95% CI 21–72), despite independent confirmation of the original sonographic diagnosis by 2 radiologist reviewers.
Sonographic evidence of vermian aplasia invariably correlated with Dandy-Walker malformation complex pathology at autopsy, but no other sonographic variable monitored in the initial blinded review distinguished the concordant and discordant groups. Although the depth of the cisterna magna tended to be greater for the concordant than for the discordant cases, this trend did not achieve statistical significance (P > .05, Mann-Whitney test). Autopsy findings that increased the likelihood of diagnostic concordance included aplasia of the vermis and cystic dilatation of the fourth ventricle.
To identify additional features that might distinguish concordant and discordant cases, the 11 cases with adequate archival ultrasound and pathology images were evaluated in an unblinded fashion. A distinct difference in the shape of the enlarged fourth ventricle was recognized, which appeared to distinguish between the groups. A representative concordant case is shown in Figure 1. A trapezoid-shaped gap is present between the cerebellar hemispheres (Fig. 1A). An in situ photograph of the cerebellum identifies similar findings at autopsy (Fig. 1B).
In contrast, a representative discordant case is shown in Figure 2. A transverse cerebellar sonogram shows a keyhole-shaped gap between the cerebellar hemispheres, with a mildly enlarged cisterna magna (Fig. 2A). Originally and during blinded review, the latter features were interpreted as Dandy-Walker malformation complex with vermian hypoplasia. However, examination of the cerebellum at autopsy, both in situ and ex situ, failed to confirm Dandy-Walker complex or any other anomalies of the posterior fossa (Figs. 2B, 2C, 2D).
The ultrasound finding of a keyhole-shaped interhemispheric gap was a consistent finding among the cases with discordant ultrasound and pathology diagnoses (5/5 cases). In contrast, the trapezoid-shaped interhemispheric gap was identified in 6/6 cases with a concordant ultrasound and pathology diagnosis of Dandy-Walker malformation complex. Available transcerebellar sonographic images from the entire series of cases were rereviewed to specifically address whether a keyhole-shaped gap between the cerebellar hemispheres was present. None of 12 cases in which there was concordance between the ultrasound and pathology diagnosis of Dandy-Walker complex had a keyhole-shaped interhemispheric gap. In contrast, a keyhole-shaped interhemispheric gap was identified in 7 of 10 cases with discordant prenatal ultrasound and pathology diagnoses of Dandy-Walker complex (95% CI 40–89; P < .001, Fisher exact test).
In this retrospective series of 44 autopsies with prenatal ultrasound diagnoses of Dandy-Walker malformation complex, including 9 with isolated Dandy-Walker malformation complex, concordance between ultrasound and neuropathological diagnosis was poor (41%, 95% CI 28–56). This major incongruity occurred despite very reasonable interobserver agreement regarding the ultrasound and autopsy interpretations. Review of the sonographic and pathologic findings suggests that a keyhole-shaped appearance of the fourth ventricle in axial ultrasound images of the inferior cerebellum may be a normal variant that is often misinterpreted as Dandy-Walker complex.
Previous reports note different rates of concordance between prenatal ultrasound findings and pathology in the diagnosis of Dandy-Walker complex (Table 1). This variability may reflect small study populations, selection biases, and/or limited clinical, radiologic, or pathologic follow-up. Other authors have cautioned about the limitations of ultrasonography in the prenatal diagnosis of Dandy-Walker malformation complex,12,21,29 but few sonographic or pathologic findings have been identified that contribute to diagnostic concordance.21,22,30
Many factors might be responsible for the correlation between ultrasonographic and pathologic diagnosis of Dandy-Walker malformation complex observed in our series and others. First, criteria used to define this complex vary. Some of these criteria (eg, dilatation of the fourth ventricle) may be easier to appreciate by ultrasound or MRI studies than by anatomical dissection. In addition, anatomical relationships that may depend on pressurized fluid in the fourth ventricle (eg, rotation of the cerebellar vermis) may be lost during dissection. Thus, variants of Dandy-Walker complex not associated with macroscopic or microscopic vermian dysgenesis may be overlooked at autopsy. In this study, we considered hypoplasia or aplasia of the vermis an obligate feature of Dandy-Walker complex, in part because limited data suggest that isolated dilatation of the fourth ventricle without associated vermian hypoplasia does not carry the same risk for functional neurologic impairment as Dandy-Walker malformation complex.8,31,32 However, long-term studies of living individuals are needed to address this issue.
Ultrasound imaging technique and interpretation may also affect the correlation between prenatal ultrasound and pathologic findings.29,33 Especially problematic is the sonographic diagnosis of agenesis of the inferior cerebellar vermis, otherwise referred to as Dandy-Walker variant or inferior vermian agenesis. For example, Ecker et al8 reported 10 cases with an ultrasound diagnosis of isolated Dandy-Walker variant, of which 6 were normal in the neonatal period. A potential cause of false-positive diagnoses of Dandy-Walker variant is scanning in an exaggerated semicoronal plane,29 which was not an issue in our series. Pilu et al33 opine that axial scans can also lead to both underestimation and overestimation of posterior fossa defects. In our series, 2 sonographic variables in axial images correlated particularly well with the pathologic diagnosis: the severity of the vermian defect and the shape of the gap between the hemispheres in transverse cerebellar images. A trapezoidal gap, with a broad dorsal base and relatively straight sides, corresponded to hypoplasia or aplasia of the vermis at autopsy. In contrast, a keyhole-shaped gap with convex sides was not associated with macroscopic or microscopic cerebellar pathology.
The anatomical basis for the keyhole-shaped gap on ultrasonography in some morphologically normal cerebella is not clear. Apparent lateral displacement of the cerebellar hemispheres may be a physiologic event that occurs normally in some fetuses, or it may represent a mild end of the Dandy-Walker malformation spectrum with no recognizable associated cerebellar pathology. One possibility is that physiologic variability in the size or position of the choroid plexus, a highly vascular tissue that normally resides along the medial aspect of the cerebellar tonsils, may influence the apparent gap between the cerebellar hemispheres (Fig. 3). Until the significance of a keyhole-shaped defect is clarified, we believe that this sonographic finding should not be regarded as a sole diagnostic criterion for Dandy-Walker malformation complex. Clinicians and their patients need to recognize that the clinical significance of an isolated keyhole-shaped defect is unknown and may not be understood until additional studies are performed that provide postnatal follow-up for infants with this prenatal finding.
To improve the diagnostic accuracy of prenatal ultrasonography in the evaluation of the posterior fossa, the use of a second scan plane has been advocated.19,20 In addition, normative data on the sonographic vertical dimensions of the fetal cerebellar vermis have been reported.34,35 The sensitivity of this measurement for predicting vermian abnormality may be low. Even with meticulous review of images, we found it difficult to determine agenesis of the inferior cerebellar vermis based solely on the sonographic appearance of the vermis. For this reason we sought ancillary findings, ie, apparent continuity between the cisterna magna and the fourth ventricle and the shape of this fluid-filled space. In the future, fetal MRI may be a useful adjunct to prenatal ultrasonography.31,36,37 In France, late second-trimester MRI has been used extensively to better characterize posterior fossa malformations initially diagnosed by ultrasonography.38
Our results indicate that prenatal sonographic diagnosis of Dandy-Walker malformation complex remains difficult and often cannot be confirmed at autopsy. We propose that the prenatal ultrasound diagnosis of Dandy-Walker malformation complex be reserved for cases with a trapezoid-shaped gap between the cerebellar hemispheres, complete vermian agenesis, and an enlarged depth of the cisterna magna. In cases that fail to meet these criteria, particularly those with only partial vermian hypoplasia and/or a keyhole-shaped gap between the cerebellar hemispheres, we recommend sequential ultrasound examinations and possible prenatal MRI. Patients need to be counseled as to the uncertain clinical significance of an isolated keyhole-shaped defect. If an affected pregnancy is terminated, a nondisruptive method of delivery of a nonautolyzed fetus should be encouraged, and comprehensive neuropathology examination should be performed. The latter should include careful gross and microscopic examination of the cerebellum, particularly the vermis, with comparable in situ and ex situ photographs that document the important anatomical relationships.
1. Sutton J. The lateral recesses of the fourth ventricle: their relation to certain cysts and tumors of the cerebellum, and to occipital meningocele. Brain 1887;9:352–61.
2. Taggert J, Walker AE. Congenital atresia of the foramens of Luschka and Magendie. Arch Neurol Psychiatr 1942;48:583–612.
3. Benda CE. The Dandy-Walker syndrome or the so-called atresia of the foramen Magendie. J Neuropathol Exp Neurol 1954;13:14–29.
4. Hirsch JF, Pierre-Kahn A, Renier D, Sainte-Rose C, Hoppe-Hirsch E. The Dandy-Walker malformation: a review of 40 cases. J Neurosurg 1984;61:515–22.
5. Hart MN, Malamud N, Ellis WG. The Dandy-Walker syndrome: a clinicopathological study based on 28 cases. Neurology 1972;22:771–80.
6. Nyberg DA, Cyr DR, Mack LA, Fitzsimmons J, Hickok D, Mahony BS. The Dandy-Walker malformation prenatal sonographic diagnosis and its clinical significance. J Ultrasound Med 1988;7:65–71.
7. Nyberg DA, Mahony BS, Hegge FN, Hickok D, Luthy DA, Kapur R. Enlarged cisterna magna and the Dandy-Walker malformation: factors associated with chromosome abnormalities. Obstet Gynecol 1991;77:436–42.
8. Ecker JL, Shipp TD, Bromley B, Benacerraf B. The sonographic diagnosis of Dandy-Walker and Dandy-Walker variant: associated findings and outcomes. Prenat Diagn 2000;20:328–32.
9. Estroff JA, Scott MR, Benacerraf BR. Dandy-Walker variant: prenatal sonographic features and clinical outcome. Radiology 1992;185:755–8.
10. Golden JA, Rorke LB, Bruce DA. Dandy-Walker syndrome and associated anomalies. Pediatr Neurosci 1987;13:38–44.
11. Ulm B, Ulm MR, Deutinger J, Bernaschek G. Dandy-Walker malformation diagnosed before 21 weeks of gestation: associated malformations and chromosomal abnormalities. Ultrasound Obstet Gynecol 1997;10:167–70.
12. Chang MC, Russell SA, Callen PW, Filly RA, Goldstein RB. Sonographic detection of inferior vermian agenesis in Dandy-Walker malformations: prognostic implications. Radiology 1994;193:765–70.
13. Russ PD, Pretorius DH, Johnson MJ. Dandy-Walker syndrome: a review of fifteen cases evaluated by prenatal sonography. Am J Obstet Gynecol 1989;161:401–6.
14. Keogan M, DeAtkine AB, Hertzberg BS. Cerebellar vermian defects: antenatal sonographic appearance and clinical significance. J Ultrasound Med 1994;13:607–11.
15. Has R, Ermis H, Yuksel A, Ibrahimoglu L, Yildirim A, Sezer HD, et al. Dandy-Walker malformation: a review of 78 cases diagnosed by prenatal sonography. Fetal Diagn Ther 2004;19:342–7.
16. Grinberg I, Northrup H, Ardinger H, Prasad C, Dobyns WB, Millen KJ. Heterozygous deletion of the linked genes ZIC1 and ZIC4 is involved in Dandy-Walker malformation. Nat Genet 2004;36:1053–5.
17. Mahony BS, Callen PW, Filly RA, Hoddick WK. The fetal cisterna magna. Radiology 1984;153:773–6.
18. Pilu G, Romero R, De Palma L, Rizzo N, Jeanty P, Copel JA, et al. Antenatal diagnosis and obstetric management of Dandy-Walker syndrome. J Reprod Med 1986;31:1017–22.
19. D'Addario V. The role of ultrasonography in recognizing the cause of fetal cerebral ventriculomegaly. J Perinat Med 2004;32:5–12.
20. Bernard JP, Moscoso G, Renier D, Ville Y. Cystic malformations of the posterior fossa. Prenat Diagn 2001;21:1064–9.
21. Carroll SG, Porter H, Abdel-Fattah S, Kyle PM, Soothill PW. Correlation of prenatal ultrasound diagnosis and pathologic findings in fetal brain abnormalities. Ultrasound Obstet Gynecol 2000;16:149–53.
22. Sun CC, Grumbach K, DeCosta DT, Meyers CM, Dungan JS. Correlation of prenatal ultrasound diagnosis and pathologic findings in fetal anomalies. Pediatr Dev Pathol 1999;2:131–42.
23. Isaksen CV, Eik-Nes SH, Blaas HG, Torp SH. Comparison of prenatal ultrasound and postmortem findings in fetuses and infants with central nervous system anomalies. Ultrasound Obstet Gynecol 1998;11:246–53.
24. Aletebi FA, Fung KF. Neurodevelopmental outcome after antenatal diagnosis of posterior fossa abnormalities. J Ultrasound Med 1999;18:683–9.
25. Kolble N, Wisser J, Kurmanavicius J, Bolthauser E, Stallmach T, Huch A, et al. Dandy-walker malformation: prenatal diagnosis and outcome. Prenat Diagn 2000;20:318–27.
26. Wald M, Lawrenz K, Deutinger J, Weninger M. Verification of anomalies of the central nervous system detected by prenatal ultrasound. Ultraschall Med 2004;25:214–7.
27. Cardoza JD, Goldstein RB, Filly RA. Exclusion of fetal ventriculomegaly with a single measurement: the width of the lateral ventricular atrium. Radiology 1988;169:711–4.
28. Benke PJ. The isotretinoin teratogen syndrome. JAMA 1984;251:3267–9.
29. Laing FC, Frates MC, Brown DL, Benson CB, DiSalvo DN, Doubilet PM. Sonography of the fetal posterior fossa: false appearance of mega-cisterna magna and Dandy-Walker variant. Radiology 1994;192:247–51.
30. Pilu G, Hobbins JC. Sonography of fetal cerebrospinal anomalies. Prenat Diagn 2002;22:321–30.
31. Klein O, Pierre-Kahn A, Boddaert N, Parisot D, Brunelle F. Dandy-Walker malformation: prenatal diagnosis and prognosis. Childs Nerv Syst 2003;19:484–9.
32. Boddaert N, Klein O, Ferguson N, Sonigo P, Parisot D, Hertz-Pannier L, et al. Intellectual prognosis of the Dandy-Walker malformation in children: the importance of vermian lobulation. Neuroradiology 2003;45:320–4.
33. Pilu G, Visentin A, Valeri B. The Dandy-Walker complex and fetal sonography. Ultrasound Obstet Gynecol 2000;16:115–7.
34. Zalel Y, Seidman DS, Brand N, Lipitz S, Achiron R. The development of the fetal vermis: an in-utero sonographic evaluation. Ultrasound Obstet Gynecol 2002;19:136–9.
35. Achiron R, Kivilevitch Z, Lipitz S, Gamzu R, Almog B Zulel Y. Development of the human fetal pons: in utero ultrasonographic study. Ultrasound Obstet Gynecol 2004;24:506–10.
36. Barkovich AJ, Kjos BO, Norman D, Edwards MS. Revised classification of posterior fossa cysts and cystlike malformations based on the results of multiplanar MR imaging. AJR Am J Roentgenol 1989;153:1289–300.
37. Levine D, Barnes PD, Madsen JR, Abott J, Mehta T, Edelman RR. Central nervous system abnormalities assessed with prenatal magnetic resonance imaging. Obstet Gynecol 1999;94:1011–9.
38. Adamsbaum C, Moutard ML, Andre C, Merzoug V, Ferey S, Quere MP, et al. MRI of the fetal posterior fossa. Pediatr Radiol 2005;35:124–40.