CSF shunts for dementia, incontinence, and gait disturbance

From a review of our experience in the past 4 years and of the literature generally, the following comments can be made about selecting patients with idiopathic NPH for a shunt procedure. (a) In the clinical presentation, either significant gait difficulty or the full triad of dementia, ataxia, and incontinence should be present. If dementia occurred first or is the major symptom, shunting may not improve the patient. (b) A CT scan with periventricular low density and/or small sulci along with expansion of the entire ventricular system (especially the temporal horns) is strongly associated with good shunt outcome. However, presence of significant atrophy does not prevent shunt success if the clinical picture is appropriate. Some surgeons now feel that the clinical presentation and CT scan findings are enough in themselves to indicate a shunt. If further testing is desired, the following may be useful: Lumbar puncture: A pressure over 100 mm is associated with better chances of improvement. Improvement after lumbar puncture is associated with high likelihood of shunt success, but lack of improvement after lumbar puncture is not useful as a predictor. Isotope or metrizamide cisternography: A typical NPH pattern suggests a good response; a mixed or normal pattern is irrelevant to shunt outcome. Overnight recording of CSF pressure: If pressure is above 180 mm at night, or if there are frequent B-waves, shunting is likely to be helpful. Lumboventricular perfusion: This technique appears to give the most accurate prediction but requires special expertise and probably human studies approval to be done, as it is still an experimental procedure. These features make it difficult to use as a routine test. With regards to results of shunting once accomplished, it is important to follow patients carefully to exclude a chronic subdural collection. If a shunted patient fails to improve with persistent large ventricles and a medium or high pressure valve was used, consideration should be given to shunt revision with insertion of a lower pressure valve.