Is there even such a thing as “idiopathic normal pressure hydrocephalus”?
There are 2 recommendations Dr Saper made in his editorial which, although based on sound scientific rationale, would be problematic to deploy. Dr Saper proposes a moratorium on all shunting procedures for NPH until evidence from a randomized, double‐blind, placebo‐controlled trial demonstrates benefits of shunting. The can of shunting has been opened and its allure would be hard to resist for a group of patients with hydrocephalus for whom few other therapeutic strategies exist. We all have followed patients whose postshunting function, by whatever mechanism (including placebo), substantially improved. Even if time limited, patients and families are likely to accept the risks of shunting for a beneficial outcome that might only last several months. Furthermore, not offering shunting to “NPH” patients because of doubts about its nature may be comparable to not offering functional neurosurgery procedures to patients with essential tremor, another disorder whose very existence is questioned.3
Dr Saper also suggests a protocol for a future randomized clinical trial; programmable‐valve shunts are implanted in the closed position, and at 1 week, patients would be blindly randomized to have the shunt turned on or not. Cognitive, gait, and continence endpoints would be completed by blinded investigators at 1 month, and then the valve switched to the opposite direction, with reevaluation after another month. The valve would then be switched on in all subjects for an open‐label phase, and subjects reevaluated at 6‐month intervals. We would argue that an early response to shunting in the valve‐open group would not exclude patients with neurodegenerative disorders who can respond to shunting temporarily. Therefore, a much longer blinded phase would be required and this may not be possible for a variety of reasons, including the impact on recruitment if patients and family members knew that the closed‐valve group would not receive active treatment for a long period of time.
A single center attempt to conduct a randomized clinical trial of ventriculoperitoneal shunting for patients with the full NPH triad failed to meet enrollment.4 A future clinical trial could only be done as a multicenter effort; if 31 cases are shunted over a 10‐year period at a single center, only 3 cases would be available per year per institution, assuming similar stringency in patient selection, which excludes a priori almost 80% of all referred cases.1 This means that at least 20 centers would be needed, assuming a sample of 60 cases to be randomized into immediate (n = 30) versus delayed shunting (n = 30) with programmable valves, which may not be conservative enough if the anticipated effect size was <0.25. Also, the follow‐up time is critical; at least 1 year postshunting in the double‐blind, valve‐open‐versus‐valve‐closed phase may be required for a pathologically heterogeneous syndrome, followed by at least 2 if not 3 years of follow‐up after the blinded period. This assumes that 1 year of “early shunting” would suffice for its disease‐modifying intervention to be statistically and clinically separable from those of a “delayed” intervention.