Hydrocephalus

Many years ago I had an article rejected from a neurology journal. One of the reviewers stated that my observation of a deleterious effect of the drug treatment of Parkinson's disease was not new, and, more importantly, might result in misunderstandings that would lead doctors to not use the drug when indicated. It would therefore be better, he opined, for doctors to not know about this problem (which evidently was known anyway). Luckily I persevered and had the article published in a more prestigious journal with a much larger readership. The comment, suggesting hiding an iatrogenic problem to "save" doctors from themselves, has, however, stayed with me.

It is therefore somewhat ironic that I am now in the position of stating something similar about another syndrome, normal pressure hydrocephalus (NPH). I think that few neurologists would differ with my opinion that many more people have been hurt than helped by being diagnosed with this condition.

NPH is generally defined as an idiopathic triad consisting of dementia, urinary incontinence and a gait abnormality. It is a rare syndrome, occurring in about one person per million people per year, which translates into one person having the syndrome each year in Rhode Island. Yet NPH is a commonly considered diagnosis in the evaluation of demented people, partly because dementing disorders are so rarely reversible that no one wants to miss one that is. Interestingly, what few doctors appreciate, is that dementia is a minor feature of this disorder, and when dementia is prominent, the patient does not have NPH.

About ten years ago I was asked to write a chapter on gait disorders for a geriatrics textbook, then in its fifth edition or so. I told the editor that I wanted to include NPH in my chapter but he responded that NPH should be in the chapter on dementia. I argued my point and two weeks later he called me to say that after he thought about his own experience with NPH over four decades at a major medical university, he couldn't recall a single case in which shunting reversed a significant dementia. NPH was included in my chapter.

The main problem with the diagnosis of NPH is that it is a syndrome. There is no pathology. The second main problem is that the literature describing it is of poor quality and contradictory. The third problem is that its triad of symptoms is very common in virtually every dementing syndrome. Later stage Alzheimer patients, vascular dementia patients and patients with dementia with Lewy bodies generally develop parkinsonian features with gait dysfunction that falls well into the highly variable descriptions of the gait abnormality of NPH, and urinary incontinence is extremely common as well. Therefore the only determinant of NPH is the imaging study. These usually show atrophy. What varies therefore is the distribution and amount of the atrophy, but in non-diagnostic fashion.

The only way to diagnose NPH is to shunt the patient. While a variety of diagnostic tests have been proposed, none have been found to be reliable. They either overlook patients who benefit from shunting, or indicate that shunting will help the patient and it doesn't. And shunting itself sometimes produces only a temporary benefit. This temporary benefit is thought to explain why the reports of shunting efficacy in the surgical literature are so much better than reports published elsewhere. The follow up time is generally much shorter.

Most fascinating is the observation that patients who have definable neurodegenerative disorders often benefit from shunting, albeit temporarily. Thus a patient I saw with two neurodegenerative disorders, progressive supranuclear palsy and dementia with Lewy bodies, diagnosed at autopsy, improved dramatically after shunting, but only for three months. Those three months were well worth the bother of the shunting. In one large study the NPH subjects underwent brain biopsy at the time of the shunting and 75% of them had Alzheimer's disease by pathological criteria. In some reports the Alzheimer patients improve more and in some less, than the non-AD patients with shunting (a problem emblematic of the morass of the literature on this syndrome). And there's a project now in progress using shunting to treat Alzheimer's disease.

The popularity of diagnosing NPH goes in cycles. It certainly has become less popular with time and experience. In Rhode Island some neurologists advocate shunting Alzheimer patients who have gait problems because, they say, their gait improves, and although this is temporary, it makes them easier to place in nursing homes. And, of course, being able to walk usually improves quality of life, although not always.

Shunting is an easy procedure. The ventricles are large and hard to miss. The problem is that the target population is elderly and frail and side effects are not insignificant. While the vast majority of the data on morbidity reflect old technology, that morbidity was in the 30% range for strokes, seizures and subdural hematomas. The new programmable shunts are believed to have reduced these problems considerably, but that's not clear yet.

When I was in training, NPH was becoming a popular diagnosis as CT scans were first being introduced, revealing that many elderly people had large ventricles. Some experts reported that one could tell that a patient had NPH if he could crawl but not walk. So patients who couldn't walk tried crawling at the neurology conferences. My chief, in some disgust at the propagation of NPH hype, suggested that maybe every demented patient should be shunted. Well, with the shunting for Alzheimer's disease, we are slowly advancing along that path.

NPH does exist. Some people do really have it and shunting can be dramatically helpful. Unfortunately then NPH is not a myth. It might be better if it was.

-JOSEPH H. FRIEDMAN, MD

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