The lack of relationship between intracranial pressure and cerebral ventricle indices based on brain computed tomography in patients undergoing ventriculoperitoneal shunt.

Abstract

In this study we investigated whether cerebral ventricle indices based on brain computed tomography (CT) scans are reliable for predicting intracranial pressure (ICP) in hydrocephalic patients. Electronic medical records of 221 patients undergoing ventriculoperitoneal shunt due to hydrocephalus were retrospectively reviewed. Cerebral ventricle indices including Evans' index, third ventricle index, cella media index, and ventricular score were calculated from transverse diameters measured at various levels on preoperative brain CT scans. ICP was considered as CSF opening pressure. Patients were categorized into three groups: communicating hydrocephalus, non-communicating hydrocephalus, and normal pressure hydrocephalus (NPH). The non-communicating hydrocephalus group was further divided according to the obstruction site; aqueduct, fourth ventricle outlet, third ventricle, and the foramen of Monro. The primary endpoint was the extent of the correlation between cerebral ventricle indices and ICP in each hydrocephalus group. No cerebral ventricle index correlated with ICP in patients with communicating hydrocephalus (n = 113) and NPH (n = 62). In the non-communicating hydrocephalus group (n = 46), only the third ventricle index revealed moderate negative correlation with ICP (r = -0.395, p < 0.01). In subgroup analyses, the third ventricle index showed a strong negative relationship with ICP only in patients with the third ventricle obstruction (r = -0.779, p < 0.05). In this study we showed that although an inverse correlation existed between ICP and the third ventricle index only in patients with non-communicating hydrocephalus due to obstruction of the third ventricle, cerebral ventricle indices based on brain CT scan were non-reliable predictors of ICP in hydrocephalic patients.