Reply

I have read Dr. Mori's letter concerning my commentary (1) on his article (2) and I wish to thank him for pointing out a mistake in a cited reference (3). Specifically, I had assumed that Dr. George's 1991 AJNR paper (3) contained information similar to his group's presentation (4) at the 1989 ASNR on large Sylvian cisterns in patients with NPH. Dr. Mori is correct in stating that large Sylvian cisterns are not mentioned as a feature of NPH in this article. On the other hand, the sign had been published by Dr. George 3 years earlier in a well-known textbook (5) in which he wrote, “Hydrocephalus, however, may coexist with large sulci and in particular with large Sylvian fissures… This occurs when the block is at the level of the high convexity or the pacchionian granulations. Consequently, the sulci and fissures dilate because of the damming of fluid proximal to the block. In effect, the sulci dilate in the same way as the ventricular system because of the distal obstruction.”

Having read this chapter and heard Dr. George's presentation, I naturally assumed that he was the first to describe the sign of large Sylvian cisterns in NPH; however, he never claimed priority. When I called him to confirm the references, he pointed out that Vassilouthis first described the sign in 1984 by using CT (6). Thus, I was correct in indicating that the sign of large Sylvian cisterns in NPH had been published previously; however, I was incorrect in my specific reference.

Concerning the issue of increased CSF flow void as an indicator of shunt-responsive NPH, I wish to point out that Dr. Mori's comments do not reflect the results as stated in either of his references, specifically in references 3 or 7. His reference 3 describes the finding of deep white matter ischemic changes with NPH (8). His reference 7 points out that the aqueductal CSF stroke volume (from quantitative phase contrast CSF velocity imaging) is increased with shunt-responsive NPH (9) not the flow void in itself.

Dr. Mori points out two articles (10, 11) that challenge the hyperdynamic CSF flow-void sign in NPH. The first of these investigators (10) used a fast spin-echo technique, and found the CSF flow void to be insensitive for predicting outcome in NPH. This is not surprising and, in fact, could have been predicted from our 1996 article in which we specifically stated that the current ubiquitous use of flow compensation minimizes the degree of signal loss attributable to CSF motion in the aqueduct (9). In that paper, 50% of the patients who were found to have hyperdynamic CSF flow by quantitative phase-contrast techniques did not have an increased flow void because the proton density–weighted images were flow-compensated. Similarly, with fast spin-echo images, the multiple 180° pulses achieve a marked degree of flow compensation, minimizing the degree of signal loss in the aqueduct and adjacent portions of the ventricular system (12). As noted (9), if flow-compensation (or fast spin-echo) techniques are used, the CSF flow void is intrinsically less marked, and if it is not seen in appropriately symptomatic patients, quantitative flow measurements should be made. The second citation (11) is an article written by two neurosurgeons. Unfortunately, they have not provided enough technical MR imaging information in their article to enable the reader to assess whether the CSF flow quantitation techniques were adequate or if the criteria for determining normal versus increased flow were appropriate.

In summary, we continue to believe firmly in the finding of hyperdynamic CSF flow as an indicator of shunt-responsive NPH—whether this is documented on proton density–weighted or conventional spin-echo images or through direct, quantitative phase-contrast CSF flow measurements. It has served us well for over 10 years of observation. Furthermore, we are grateful to Dr. Mori for making the finding of large Sylvian cisterns with NPH more accessible in the peer-reviewed radiologic literature.