The management of obstructive hydrocephalus associated to posterior cranial fossa tumor remains a controversial issue. In this context, the role of endoscopic third ventriculostomy (ETV) performed before tumor removal (the so-called preoperative ETV) is debated and its success rate is reported as being largely variable in the literature. We describe two children who experienced the closure of a preoperatively performed ETV after the surgical resection of their posterior fossa tumor. In both cases tumor removal was performed in the prone position. Radiological investigations and intraoperative evidence documented that the failure of the ETV was due to a blood clot which directly occluded the stoma. We speculate that the prone position favored the migration of blood, through the aqueduct, thereafter, clotting on the most declivous part of the third ventricle once the patient assumed the upright position, with the consequent occlusion of the stoma. Based on a gravitational hypothesis, we named this mechanism the “snow globe effect.”

Introduction: Intraoperative ultrasound (IOUS) may aid the resection of space-occupying brain lesions, though technical limits may hinder its reliability. Methods: IOUS (MyLabTwice^®, Esaote, Italy) with a microconvex probe was utilized in 45 consecutive cases of children with supratentorial space-occupying lesions aiming to localize the lesion (pre-IOUS) and evaluate the extent of resection (EOR, post-IOUS). Technical limits were carefully assessed, and strategies to enhance the reliability of real-time imaging were accordingly proposed. Results: Pre-IOUS allowed us to localize the lesion accurately in all of the cases (16 low-grade gliomas, 12 high-grade gliomas, eight gangliogliomas, seven dysembryoplastic neuroepithelial tumors, five cavernomas, and five other lesions, namely two focal cortical dysplasias, one meningioma, one subependymal giant cell astrocytoma, and one histiocytosis). In 10 deeply located lesions, IOUS with hyperechoic marker, eventually coupled with neuronavigation, was useful to plan the surgical route. In seven cases, the administration of contrast ensured a better definition of the vascular pattern of the tumor. Post-IOUS allowed the evaluation of EOR reliably in small lesions (2 cm) assessing EOR is hindered by the collapsed surgical cavity, especially when the ventricular system is opened, and by artifacts that may simulate or hide residual tumors. The main strategies to overcome the former limit are inflation of the surgical cavity through pressure irrigation while insonating, and closure of the ventricular opening with Gelfoam before insonating. The strategies to overcome the latter are avoiding the use of hemostatic agents before IOUS and insonating through normal adjacent brain instead of corticotomy. These technical nuances enhanced the reliability of post-IOUS, with a total concordance to postoperative MRI. Indeed, the surgical plan was changed in about 30% of cases, as IOUS showed a residual tumor that was left behind. Conclusion: IOUS ensures reliable real-time imaging in the surgery of space-occupying brain lesions. Limits may be overcome with technical nuances and proper training.