Hydrocephalus

Hakim and Adams first described the syndrome of Idiopathic Normal Pressure Hydrocephalus (INPH) in 1965.1965 In two articles they reported six patients with a triad of "mild impairment of the memory, slowness and paucity of thought and action, unsteadiness of gait and unwitting urinary incontinence." In three out of the six patients the etiology was unknown, while head trauma (2) and a cyst of the third ventricle accounted for the other three patients.

Their observation was based on a syndrome, known at the time, of secondary hydrocephalus. In this disorder, hydrocephalus developed as a result of blood or inflammatory cells in the cerebrospinal fluid (CSF), blocking reabsorption. With expanding hydrocephalus due to increased pressure, patients would usually develop impaired consciousness, headache, and subsequent gait abnormality and bladder incontinence. Therefore, what was new was the normal CSF pressure and the idiopathic nature of the disease.

In 1982, CM Fisher underscored the importance of the gait disturbance in the description of NPH. 3 He described 30 patients over 60 years of age, evaluated personally, who presented primarily with gait disturbance, hydrocephalus on CT scan and responsiveness to CSF drainage, either by shunting or by spinal puncture. None of the patients had dementia preceding the gait abnormality, although both problems could be present. He reported that the onset of gait abnormality before cognitive decline predicted a better prognosis after shunting.

In 1986, Graff-Radford and Godersky correlated clinical symptoms and shunt responsiveness. 4 In their series of 21 patients shunted for INPH, 16 patients improved. Of these, 11 had a gait abnormality that preceded the dementia, while 5 had a simultaneous presentation. Of the 5 patients who did not improve, 3 had a dementing illness prior to a gait abnormality, 2 had a concomitant presentation and in only one did the gait disturbance precede the dementia. Based on those findings, the authors concluded that gait abnormality preceding the dementia was a good prognostic factor in the selection of patients for shunting.

INPH is often considered one of the rare reversible causes of dementia, which is why it is so commonly discussed. However, its importance as a potential cause of reversible dementia is debatable. In a recent meta-analysis of 39 articles addressing reversibility in dementia, 7042 patients with cognitive impairment (5620 with dementia) were assessed. ^ Potential reversibility was found in 9% of the patients, of whom only 0.6% of the patients actually reversed. In 2.2% of these patients a surgically amenable condition was found, including NPH (idiopathic or secondary), subdural hematomas and cerebral tumors. According to this analysis, the estimated incidence of shunt-responsive NPH is 2.2 per million per year.

Similar figures can be derived from memory clinic reports.6 In a study of 71 consecutive patients sent for evaluation of possible INPH, 36 patients did not meet criteria based on histoiy, physical and neurological examination and neuroimaging. The most common diagnosis in the excluded population was cerebrovascular disease. Thirty-five patients were admitted to the neurosurgery ward for hydrodynamic studies, of whom 13 were considered shunt candidates. Only 8 patients respond to shunting. Considering shunt responsiveness as the only criterion to confirm the diagnosis of INPH, the calculated incidence of INPH in this population was 1.5 per million per year.

The pathophysiology of INPH is unclear, and many theories have been proposed. It is important to differentiate primary from secondary cases of NPH. The latter usually develops after meningitis, subarachnoid hemorrhage or head trauma. Secondary cases have a much better prognosis and studies evaluating NPH should differentiate between those two populations.7

Although isolated intracranial pressure (ICP) measurements in INPH are within the normal range, continuous ICP monitoring demonstrates an increase in the occurrence of B-waves in these patients. 8 While B-waves are a physiological phenomena of unclear etiology, their presence in 50% or more of the recording time in NPH patients predicts a good shunting response according to some studies.

The association between INPH and hypertension, subcortical and periventricular white matter abnormalities (WMD) suggests a possible cerebrovascular role in the development of INPH.9 The presence of WMD does not preclude shunting, however it carries a poorer prognosis.10

Abnormalities in the cerebral blood flow (CBF) in the periventricular areas and subcortical white areas, mostly in the frontal lobes have been demonstrated in different studies utilizing multiple techniques.11 Some experts believe the alterations in CBF, also known as "misery perfusion", are responsible for the symptoms. Some reports have suggested reversibility or a decrease in those abnormalities after successful shunting or CSF drainage. Unfortunately, the different techniques used in the different studies, may explain the variability of the results.

Pathological changes at the arachnoid level were described as a potential culprit in the abnormalities in CSF flow, secondary to deficits in absorption. Biopsy studies do not corroborate this view. Pathological abnormalities in the arachnoid, when observed, do not correlate with results of CSF hydrodynamic studies or shunt responsiveness. 12 In addition, deficits in absorption at the arachnoid level would not create a pressure gradient between convexities and ventricles, and would therefore not lead to hydrocephalus.7

Any doctor with geriatric experience knows that the typical "clinical triad" is a relative common finding in the elderly. Many other disease processes, either alone or in combination, may produce a similar clinical picture. Some characteristics, as described below, are considered helpful in the differential diagnosis, although there is no pathognomonic clinical or imaging sign.

The gait abnormality in INPH is of "frontal dysequilibrium" or frontal gait disorder type. The most common symptoms described by patients are imbalance, tiredness of legs, leg weakness and, as the disease progresses, shorter steps, shuffling, scuffing and slow turning. U3 A triad of hypokinetic gait, associated with a wide base, decreased step height and disturbance of dynamic equilibrium is considered "characteristic" of INPH. 13

The differentiation between INPH and Parkinsons disease can be challenging. According to one study, in INPH the most helpful gait parameters are the wide-based gait, with outward feet rotation, decreased range of motion of the lower extremities, with little or no reduction of arm swing. In contrast, PD usually has associated extrapyramidal features and the patients respond to visual and acoustic cues. 13

Cortical features such as apraxia, agnosia and aphasia are absent in INPH. Problems with attention, concentration, forgetfulness, apathy and mental slowing are the rule. Neuropsychological tests, although helpful in characterizing the deficit, are not good predictors of shunting outcome, and in most studies the improvement in extended neuropsychological batteries is minimal. 14

Iddon et al demonstrated the unpredictability of the cognitive response to shunt. 15 In this study they evaluated 11 patients with INPH, 5 in the demented range (mini-mental state examination-MMSE24. After shunting, the demented patients demonstrated significant improvement in their MMSE and the Kendrick Object Learning Test (KOLT) which assesses recall of everyday objects after a brief presentation period of a black-and-white illustration of them scores. On the other hand, the nondemented patients demonstrated decreased verbal fluency, impaired spatial recognition and attentional set shifting task. Even though the subjects demonstrated improvement of gait after shunting, the impairment on tests sensitive to frontal dysfunction remained.

Urinary urgency is common in the earlier stages; incontinence is a later sign. 7 Urodynamic studies demonstrate hyperreflexia and instability of the bladder detrusor muscle, but there is no evidence of defective bladder control. This suggests damage in the periventricular pathways to the sacral bladder center and decreased inhibition of bladder contractions.

The diagnosis of INPH relies on clinical criteria, neuro-imaging evidence of hydrocephalus and ideally evidence of abnormalities in CSF hydrodynamics. Vanneste et al evaluated the predictive value of clinical criteria and CT in the diagnosis of INPH. 16 By using strictly clinical (predominant gait disturbance, absent to mild cognitive deficit) and CT criteria (rounded frontal horns, moderate ventriculomegaly, absence of white matter disease and cortical atrophy), the positive predictive value (PPV) was 65%, and the negative predictive value (NPV) was 76%. Ineffective shunting was observed in 11% of the patients, while missed improvement occurred in 13%. More liberal criteria would have resulted in a PPV of 58%, ineffective shunting in 37% and missed improvement in 5% of the patients. The authors concluded that clinical criteria associated with appropriate neuroimaging is the standard by which new diagnostic testing should be used to improve diagnostic accuracy and shunt responsiveness.

CSF drainage by lumbar puncture, often considered a "gold standard" in the diagnosis of INPH, has a very high incidence of false negatives; and a prospective study failed to provide any significant additional information compared to conservative clinical and neuroimaging criteria were utilized to select patients for shunting. 17 Continuous CSF drainage poses a similar problem. In addition, it requires hospital admission, technical expertise, and is not always benign. 18 Cysternography does not improve the diagnostic accuracy of clinical and CT criteria combined, although many doctors still use it. 19 Continuous intracranial pressure monitoring is a useful procedure and results have been uniform. Strong evidence supports that the frequent occurrence of B-waves (>50% of the recording time) in patients with INPH predicts a good response to shunting. 7 Unfortunately continuous ICP monitoring is not a routine procedure in most neurosurgical centers.

There are many other diagnostic modalities, such as the infusion test, the conductance test, cerebrovascular and metabolic imaging studies. None has consistently demonstrated a significant improvement in predicting shunt-responsiveness.7

The differential diagnosis of NPH includes Alzheimer's disease and related dementias, various causes of Parkinsonism, most notably Parkinsons disease, dementia with Lewy bodies (DLB) and "multi-infarct encephalopathy". Before considering the diagnosis of INPH, one should exclude other possible explanations rather than the other way around. As mentioned before, INPH, is a relatively rare disease, with an incidence comparable to Creutzfeld-Jacob disease, while Alzheimer's disease, Parkinsons disease, DLB and multi-infarct state are common pathologies.

The predictors of a positive shunt response are gait predominant disorder that precedes any cognitive changes, a short history or absence of cognitive impairment, a known cause for the development of the hydrocephalus, absence of white matter lesion in neuroimaging studies. When available, the presence of B-waves for at least 50% of the recording time, improvement of symptoms after CSF drainage and CSF outflow resistance over 1 SmmHG/ ml per minute are also valuable prognostic factors. In contrast, severe dementia, dementia as the first symptom and severe atrophy or white matter disease on the MRl are predictors of a negative response to shunting7.

One may argue that it is appropriate to shunt a demented, gait-impaired patient, to improve the gait, even though no significant improvement will be expected in cognitive function. In that case, the family should have realistic expectations, and be informed of the risks of the shunt. The mean chance of significant improvement after shunting is 30 to 40% in INPH and 50-70% in secondary cases. Complications range between 20-40% of the cases, and include strokes, subdural hematomas, shunt malfunctioning, seizures and problems related to the anesthesia. Serious complications (death, severe residual deficit) usually arise in patients with severe co-morbidities, and are observed in less than 8% of the patients with INPH.7,20 Dementia is the least likely symptom to improve especially if severe.

In conclusion, INPH is an overestimated, uncommon disorder, with a common clinical triad in the elderly population. Potential improvement (but probably not reversibility) of the gait more than the cognitive dysfunction can be expected only in highly selected patients.

REFERENCES

1. Hakim S, Adams RO. J Neurol Sci 1965;2:307-27.

2. Adam RD, Fisher CM, Hakim S, et al. NEJM 1965;273:117-26.

3. Fisher CM. Neurol 1982;32:358-63.

4. Graff-Radford NR, Godersky JC. Arch Neurol 1986;43:940-2.

5. Clarfield M. Arch Intern Med 2003;163:2219-29.

6. Bech-Azeddine R, Waldemar G, Knudsen GM, et al. Eur J Neurol 2001;6:279-88.

7. Vanneste JAL. J Neurol 2000;247:5-14.

8. Symon L, Dorsch NW. Lancet 1972;2:1291-2.

9. Krauss JK, Regel JP, Werner V, et al. Stroke 1996;27:24-9.

10. Krauss JK, Regel JP, Vach W, et al. Neurosurg 1997;40:491-6.

11. Owler BK, Pickard JD. Acta Neurol Scand 2001;104:325-42.

12. Bech RA, Juhler M, Waldemar G, et al. Neurosurg 1997;40:497-502.

13. Stoltze H, Kuhtz-Buschbeck JP, Drücke H, et al. J Neurol Neurosurg Psychiatry 2001;70:289-97.

14. Savolainen S, Hurskainen H, Paljärvi L, et al. Acta Neurochir 2002;144:515-23.

15. Iddon JL, Pickard JD, Cross JL, et al. J Neurol Neurosurg Psychiatry 1999;67:723-32.

16. Vanneste J, Augustijn P, Tan WF, Dirven C. J Neurol Neurosurg Psychiatry 1993;56:251-6.

17. Malm J, Kristensen B, Karlsson T, et al. Arch Neurol 1995;52:783-9.

18. Walchenbach R, Geiger E, Thomeer RTWM, Vanneste JAL. J Neurl Neurosurg Psychiatry 2002;72:503-6.

19. Vanneste J, Augustijn P, Gareth AG, et al. Arch Neurol 1992;49:366-70.

20. Vanneste J, Augustin P, Dirven C, et al. Neurol 1992;42:54-9.

ANELYSSA D'ABREAU, MD

Anelyssa D'Abreau, MD, a former fellow in geriatric neurology at Brown Medical School/Memorial Hospital of RI, is a geriatric neurologist at the State University of Campinas (UNICAMP), Brazil.

CORRESPONDENCE:

Anelyssa D'Abreu, MD

e-mail: anelyssa@hotmail.com

Copyright Rhode Island Medical Society Dec 2004
Provided by ProQuest Information and Learning Company. All rights Reserved