Marked cerebrospinal fluid void: Indicator of successful shunt in patients with suspected normal-pressure hydrocephalus

William G. Bradley, Anthony R. Whittemore, Keith E. Kortman, Arthur S. Watanabe, Michelle Homyak, Louis M. Teresi, Stephen J. Davis

Research output: Contribution to journalArticlepeer-review

182 Scopus citations

Abstract

The authors blindly reviewed the charts of 20 patients with normal-pressure hydrocephalus (a disease of unknown cause characterized radiologically as chronic communicating hydrocephalus and clinically by gait apraxia, dementia, and incontinence) who had undergone creation of a ventriculoperitoneal shunt. The initial clinical response to surgery was graded excellent, good, fair, or poor; 5-year follow-up was available in 55% of cases. The magnetic resonance (MR) images obtained in these patients were also blindly reviewed for the magnitude of cerebrospinal fluid (CSF) flow void (graded on the basis of extent rather than degree of signal loss) in the cerebral aqueduct. A significant (P < .003) correlation existed between good or excellent response to surgery and an increased CSF flow void. The presence of associated deep white matter infarction on MR images did not correlate with a poor response to surgery. On the basis of these findings, it is suggested that patients who fulfill the clinical criteria of NPH and have an increased CSF flow void undergo creation of a shunt.

Original languageEnglish (US)
Pages (from-to)459-466
Number of pages8
JournalRADIOLOGY
Volume178
Issue number2
StatePublished - Feb 1991

Keywords

  • Brain, hydrocephalus, 10.823
  • Brain, white matter, 13.87
  • Cerebrospinal fluid, MR studies, 10.1214
  • Cerebrospinal fluid, flow dynamics, 10.1214
  • Shunts, ventriculoperitoneal, 10.4513

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Marked cerebrospinal fluid void: Indicator of successful shunt in patients with suspected normal-pressure hydrocephalus'. Together they form a unique fingerprint.

Cite this