Ultrasonographic differential diagnosis of fetal intracranial interhemispheric cysts

Ultrasonographic differentiation between intracranial supratentorial interhemispheric pathologic cystlike lesions and those related to physiologic median structures is essential because the latter have no clinical relevance, whereas the former may carry a poor prognosis. We reviewed our experience with 19 consecutive cases of interhemispheric hypoechoic lesions without parenchymal involvement diagnosed between January 1990 and June 1997 to establish their clinical significance and provide prenatal ultrasonographic criteria to distinguish between pathologic cystlike lesions and those related to physiologic midline structures.

All patients underwent targeted prenatal scans of intracranial anatomy to establish the relationship between the fluid collections and the surrounding parenchymal and ventricular structures. In addition, a detailed anatomic survey was performed to rule out associated malformations. Follow-up, including neurologic examination, imaging, autopsy evaluation, or a combination was performed in all cases. Statistical analysis used the Wilcoxon rank sum test, the Fisher exact test, and the chi2 test for trend. P <.05 was considered significant.

Cystlike lesions related to physiologic median structures (n = 12) included enlargement of the cavum septi pellucidi (n = 3), enlargement of the cavum vergae (n = 2), and cysts of the velum interpositum (n = 7). These lesions were unilocular and had a median size of 10 mm (range 10-30 mm); they resolved in 5 cases and remained stable in the remainder. They were not associated with overt abnormalities, other than borderline ventriculomegaly in 2 cases. Pediatric follow-up (median 26 months, range 3-84 months) showed normal neurodevelopment in all cases. Pathologic cystlike lesions (n = 7) were significantly larger (median 40 mm, range 10-80 mm, P =.004) and had a significantly worsening trend, growing more at serial prenatal ultrasonographic examinations (P =.039) than fluid collections related to physiologic median structures. Moreover, prenatal ultrasonographic evidence of associated intracranial abnormalities, in the form of partial or total agenesis of the corpus callosum and overt hydrocephalus, was present in 5 of 7 cases of pathologic cystlike lesions and in none of the 12 related to physiologic structures (P =.002). Median gestational age at diagnosis was not different between those with cystlike lesions related to physiologic median structures and those with pathologic lesions (30 and 31 weeks, respectively). Among the latter group, 1 pregnancy was voluntarily terminated, 1 infant died at 4 months of age, 2 infants had neurodevelopmental delay, and 3 infants were neurologically healthy at a mean follow-up of 43 months. Cyst shunting was necessary in 5 of 6 cases.

Interhemispheric cystlike lesions related to physiologic structures can be prenatally distinguished from pathologic fluid collections on the basis of location, cyst size, change in size with time, and absence of associated anomalies.