Impaired pulsation absorber mechanism in idiopathic normal pressure hydrocephalus

Abstract

The pathophysiology of normal pressure hydrocephalus (NPH), and the related problem of patient selection for treatment of this condition, have been of great interest since the description of this seemingly paradoxical condition nearly 50 years ago. Recently, Eide has reported that measurements of the amplitude of the intracranial pressure (ICP) can both positively and negatively predict response to CSF shunting. Specifically, the fraction of time spent in a "high amplitude" (> 4 mm Hg) state predicted response to shunting, which may represent a marker for hydrocephalic pathophysiology. Increased ICP amplitude might suggest decreased brain compliance, meaning a static measure of a pressure-volume ratio. Recent studies of canine data have shown that the brain compliance can be described as a frequency-dependent function. The normal canine brain seems to show enhanced ability to absorb the pulsations around the heart rate, quantified as a cardiac pulsation absorbance (CPA), with properties like a notch filter in engineering. This frequency dependence of the function is diminished with development of hydrocephalus in dogs. In this pilot study, the authors sought to determine whether frequency dependence could be observed in humans, and whether the frequency dependence would be any different in epochs with high ICP amplitude compared with epochs of low ICP amplitude. Systems analysis was applied to arterial blood pressure (ABP) and ICP waveforms recorded from 10 patients undergoing evaluations of idiopathic NPH to calculate a time-varying transfer function that reveals frequency dependence and CPA, the measure of frequency-dependent compliance previously used in animal experiments. The ICP amplitude was also calculated in the same samples, so that epochs with high (> 4 mm Hg) versus low (≤ 4 mm Hg) amplitude could be compared in CPA and transfer functions. Transfer function analysis for the more "normal" epochs with low amplitude exhibits a dip or notch in the physiological frequency range of the heart rate, confirming in humans the pulsation absorber phenomenon previously observed in canine studies. Under high amplitude, however, the dip in the transfer function is absent. An inverse relationship between CPA index and ICP amplitude is evident and statistically significant. Thus, elevated ICP amplitude indicates decreased performance of the human pulsation absorber. The results suggest that the human intracranial system shows frequency dependence as seen in animal experiments. There is an inverse relationship between CPA index and ICP amplitude, indicating that higher amplitudes may occur with a reduced performance of the pulsation absorber. Our findings show that frequency dependence can be observed in humans and imply that reduced frequency-dependent compliance may be responsible for elevated ICP amplitude observed in patients who respond to CSF shunting.