Vein of Galen malformation: correlation of clinical presentation, arteriography, and MR imaging.

To test the value of currently proposed angiographic categorizations of vein of Galen malformations and a hypothesis regarding the causes of vein of Galen malformations and of accompanying hydrocephalus, as well as to assess the relative utility of MR imaging and CT in clinical evaluation, we reviewed the clinical and radiologic records of 34 patients with vein of Galen malformations. Patients were divided into two groups on the basis of the angiographic demonstration of either an arteriovenous malformation nidus or a direct arteriovenous fistula to the wall of the vein of Galen or one of its tributaries. Patients with such a nidus (n = 17) could be distinguished from those with arteriovenous fistulas alone (n = 17) on the basis of age at presentation (p less than .01) and presenting symptoms. Venous constraints, thought to be etiologically important, were identified in 31 of 34 patients. The presence or absence of hydrocephalus was explainable by mass effect in only 24 of 32 patients. In seven of 32 cases, no obvious mass effect was seen in the presence of hydrocephalus, but arteriographic evidence of venous hypertension was present in all patients with hydrocephalus. MR provided improved depiction of both arterial and venous anatomy as compared with CT. Parenchymal abnormalities were uncommon. No patients had subarachnoid hemorrhages. We conclude that MR is superior to CT in the clinical evaluation of vein of Galen malformations, that the angiographic finding of a nidus separates patients with vein of Galen malformations into clinical and therapeutically relevant groups, and that simple mass effect on the aqueduct is not an adequate explanation for all cases of hydrocephalus in patients with this disease.