I. Normal CSF physiology

  1. Produced by choroid plexus & brain parenchyma
  2. Flows from lateral ventricles through foramen of Monroe to 3rd ventricle, then through aqueduct of Sylvius to 4th ventricle, then through foramina of Magendie and Luschka into subarachnoid space.
  3. Ultimately absorbed by arachnoid villi & through open channels around cranial & spinal nerves.

II. Pathophysiology of NPH

  1. Usually the problem seems to be flow of CSF from basal cisterns to sagittal & other sinuses, i.e. a communicating hydrocephalus
  2. Ventriculomegaly without expansion of subarachnoid space
  3. Idiopathic NPH (more common)
  1. Associated with risk factors for atherosclerosis
  1. NPH pts 3-6 times more likely than age- and sex-matched controls to have HTN, low HDL, DM, CAD, and MR evidence of deep white-matter infarction
  1. Secondary NPH
  1. Subarachnoid hemorrhage
  1. 10-20% will get symptomatic NPH
  2. Usually from acute obstruction of aqueduct of Sylvius or 4th ventricle outlets
  3. Can occur months after hemorrhage
  1. Trauma
  1. After 29-72% of head injuries
  2. Prob. due to scarring of basal subarachnoid cisterns
  1. Infection
  1. Meningitis can scar basal subarachnoid cisterns & arachnoid granulations
  2. Tends to occur weeks after meningitis
  1. Tumors obstructing any part of CSF pathway
  1. Carcinomatous meningitis can cause NPH by impeding CSF reabsorption
  1. Link between pathophysiology and clinical presentation is unclear

III. Clinical Presentation

  1. Gait abnormality
  1. Seems to associated with increased extensor tone
  2. Slow, broad-based shuffling; often described as a "gait apraxia"
  1. Dementia
  1. Memory loss
  2. Slowing of thinking
  3. Usually no sz, aphasia, or agnosia
  1. Urinary incontinence; rarely can get fecal incontinence as well

IV. Diagnosis

  1. CT: ventriculomegaly, often periventricular lucencies (esp. frontal), usually nl sized subarachnoid space, i.e. sulci aren't abnormally deep as in Alzheimer's; however, can have NPH with good response to shunting and big subarachnoid space.
  2. On MRI, can see upward bowing of corpus callosum
  3. MRI CSF flow analyses can show an abnormally great flow void near the aqueduct of Sylvius, reflecting a hyperdynamic CSF flow state
  4. Isotope cisternography: commonly used in past, not very useful now; does not add accuracy in predicting favorable response to shunting compared with CT & clinical evaluation
  5. If LP is done, CSF pressure should be nl, i.e. <18cm H20

V. Treatment

  1. Treatment of choice is shunting, usually ventriculoperitoneal, though lumboperitoneal is also done
  1. Predictors of poor response to shunting
  1. Large-vessel atherosclerotic disease
  1. Predictors of good response to shunting
  1. Good response after LP
  2. Known cause of NPH (80-98% response rate, c/w 60-74% with idiopathic NPH)
  3. Prominence of gait symptomatology
  4. Expansion of normal flow void on MRI (see above)
  1. Risks of shunt placement
  1. Occur in 35-52% for pts with idiopathic NPH
  2. Include malfunction, infection, overdrainage & resultant subdural hematoma, epilepsy, and rarely, intracerebral hemorrhage
  1. LP with or without placement of a CSF drain can be temporarily helpful
  2. Acetazolamide 250-500/d decreases CSF production and seemed helpful in one small (n = 15) uncontrolled study

(Source: Neurosurg. Clin. N. Am., 4:667, 1993)