OBJECTIVE: We investigate the predictive value of cerebrospinal fluid (CSF) flow void on outcome after shunting in a prospective series of patients with idiopathic normal pressure hydrocephalus (NPH).
METHODS: The degree and extension of CSF flow void were examined on T2-weighted magnetic resonance imaging scans of 37 elderly patients with idiopathic NPH who underwent subsequent shunting. The degree of flow void was assessed in comparison with the signal of large cerebral arteries. The extension was evaluated via the calculation of sum scores for the occurrence of flow void in different locations of the ventricular system. Those parameters were not considered in the decision to perform shunting. CSF flow void in the aqueduct and the adjacent third and fourth ventricles of the 37 patients with idiopathic NPH was compared with that of 37 age-matched control patients. CSF flow void scores in patients with idiopathic NPH were investigated for correlations between postoperative outcome scores and ventricular width indices.
RESULTS: No difference was found between the occurrence of aqueductal CSF flow void in patients with idiopathic NPH and the control group. A significant difference, however, was noted for the extension of the CSF flow void, which was greater in the NPH group. Postoperative improvement was found in 33 of 37 patients with idiopathic NPH at a mean follow-up of 15.6 months. Only small, statistically not significant correlations were found between CSF flow void and postoperative outcome. Flow void sum scores, however, correlated significantly with ventricular width indices.
CONCLUSION: The degree and extension of CSF flow void on T2-weighted magnetic resonance imaging scans have little predictive value for outcome after shunting in patients with idiopathic NPH. The greater extension of the CSF flow void in patients with NPH is most likely related to increased ventricular width. It is not useful to consider CSF flow void findings on conventional magnetic resonance imaging scans in making the decision to offer shunting in patients with idiopathic NPH.
Departments of Neurosurgery (JKK, JPR) and Neurology (DWD), Section of Neuroradiology (FDJ, AKW), Institute of Medical Biometry and Medical Informatics (WV), and Center for Data Analysis and Model Building (WV), Albert-Ludwigs-Universität, Freiburg, Germany
Received, April 9, 1996. Accepted, July 31, 1996.
Reprint requests: Priv. Doz. Dr. Joachim K. Krauss, Baylor College of Medicine, Department of Neurosurgery, 6560 Fannin, Suite 944, Houston, Texas 77030.