Dandy-Walker malformation complex: correlation between ultrasonographic diagnosis and postmortem neuropathology.

OBJECTIVE: This autopsy-based study was designed to evaluate sonographic and neuropathologic findings of fetuses diagnosed prenatally with Dandy-Walker malformation complex. METHODS: The retrospective study encompassed a series of 44 autopsy cases from 2 tertiary referral centers with a prenatal ultrasound diagnosis of Dandy-Walker malformation complex between 1995 and 2003. Ultrasound and pathology data from the cases and from age-matched controls were reviewed in a blinded manner. An unequivocal diagnosis of Dandy-Walker malformation complex from ultrasonography or pathology archival images required significant hypoplasia or aplasia of the cerebellar vermis. RESULTS: Neuropathologic examination failed to confirm the prenatal diagnosis of Dandy-Walker malformation complex in 59% (26/44, 95% confidence interval [Cl] 44-72) of the cases. After standardized reevaluation of high quality archival sonograms and pathology images, concordance remained poor at 55% (6/11 cases, 95% Cl 28-79). Sonographic features that favored concordance included marked enlargement of the cisterna magna (≥ 10 mm), complete aplasia of the vermis, and a trapezoid-shaped gap between the cerebellar hemispheres. This latter finding contrasted with a keyhole-shaped gap in fetuses with no cerebellar neuropathology. CONCLUSION: Correlation between a prenatal ultrasound diagnosis of Dandy-Walker malformation complex and autopsy neuropathology findings is poor. Unequivocal prenatal sonographic diagnosis of Dandy-Walker malformation complex should be reserved for cases with the classic findings of Dandy-Walker malformation, including enlargement of the cisterna magna, aplasia of the vermis, and a trapezoid-shaped, rather than keyhole-shaped, interhemispheric gap.