Cerebrospinal fluid pulse pressure amplitude during lumbar infusion in idiopathic normal pressure hydrocephalus can predict response to shunting

Abstract

Background: We have previously seen that idiopathic normal pressure hydrocephalus (iNPH) patients having elevated intracranial pressure (ICP) pulse amplitude consistently respond to shunt surgery. In this study we explored how the cerebrospinal fluid pressure (CSFP) pulse amplitude determined during lumbar infusion testing, correlates with ICP pulse amplitude determined during over-night ICP monitoring and with response to shunt surgery. Our goal was to establish a more reliable screening procedure for selecting iNPH patients for shunt surgery using lumbar intrathecal infusion.

Methods: The study population consisted of all iNPH patients undergoing both diagnostic lumbar infusion testing and continuous over-night ICP monitoring during the period 2002-2007. The severity of iNPH was assessed using our NPH grading scale before surgery and 12 months after shunting. The CSFP pulse was characterized from the amplitude of single pressure waves.

Results: Totally 62 iNPH patients were included, 45 of them underwent shunt surgery, in whom 78% were shunt responders. Among the 45 shunted patients, resistance to CSF outflow (R(out)) was elevated (>or= 12 mmHg/ml/min) in 44. The ICP pulse amplitude recorded over-night was elevated (i.e. mean ICP wave amplitude >or= 4 mmHg) in 68% of patients; 92% of these were shunt responders. In those with elevated overnight ICP pulse amplitude, we found also elevated CSFP pulse amplitude recorded during lumbar infusion testing, both during the opening phase following lumbar puncture and during a standardized period of lumbar infusion (15 ml Ringer over 10 min). The clinical response to shunting after 1 year strongly associated with the over-night ICP pulse amplitude, and also with the pulsatile CSFP during the period of lumbar infusion. Elevated CSFP pulse amplitude during lumbar infusion thus predicted shunt response with sensitivity of 88 and specificity of 60 (positive and negative predictive values of 89 and 60, respectively).

Conclusions: In iNPH patients, shunt response can be anticipated in 9/10 patients with elevated overnight ICP pulse amplitude, while in only 1/10 with low ICP pulse amplitude. Additionally, the CSFP pulse amplitude during lumbar infusion testing was elevated in patients with elevated over-night ICP pulse amplitude. In particular, measurement of CSFP pulse amplitude during a standardized infusion of 15 ml Ringer over 10 min was useful in predicting response to shunt surgery and can be used as a screening procedure for selection of iNPH patients for shunting.

Figure 1

Analysis of lumbar infusion tests: (a) a plot showing our conventional method of performing a lumbar infusion test, using constant-rate infusion at 1.5 ml/min of Ringer solution. The plot of mean CSFP against time incorporates the period before infusion (opening pressure; P_o), and during the infusion period (infusion rate of 1.5 ml/min). The plateau pressure (P_p) is indicated. The resistance to CSF outflow (Rout) is calculated as (P_p - P_o)/infusion rate. Two 6 s time windows of the continuous CSFP signal from (b) the start and (c) the termination of the infusion test are indicated, illustrating the single CSFP pulse waves. (d) Plot of mean CSFP pulse amplitude against time, indicating the period after lumbar puncture and before infusion (opening phase), and the 10 min period during the infusion. For comparison between patients we compared the CSFP pulse amplitude during the first 10 min of infusion.

Figure 2

Scatter plots of CSFP wave amplitude during the opening period and during lumbar infusion for the two categories of over-night pulsatile ICP (mean ICP wave amplitude either < 4 mmHg or ≥ 4 mmHg during the period 11 p.m. to 7 a.m.) a) Shows the pulsatile CSFP (mean CSFP wave amplitude) during the opening phase after lumbar puncture, and (b) during the phase of lumbar infusion (15 ml over 10 min; 1.5 ml/min). Dotted lines denote different targets for CSFP wave amplitude during the opening period and during lumbar infusion. Significant differences between groups were tested with one-way ANOVA.

Figure 3

Graphs showing the correlation between changes in NPH Score one year after shunt surgery and (a) ICP wave amplitude during over-night ICP monitoring, (b) CSFP wave amplitude during the opening phase after lumbar puncture, and (c) CSFP wave amplitude during the phase of lumbar infusion (15 ml over 10 min; 1.5 ml/min).

Figure 4

Scatter plots of mean CSFP wave amplitude for shunt response categories non-responder and responder during (a) over-night ICP monitoring, (b) during the opening phase after lumbar puncture, and (c) during the phase of lumbar infusion. Dotted lines denote cut off targets for the pulsatile pressure (wave amplitude). The pulse amplitude during the opening phase of lumbar puncture was not as good a predictor for shunt response as the amplitude during overnight monitoring and during lumbar infusion. Significant differences between responders and non-responders are shown (one-way ANOVA).

Figure 5

Assessments of intracranial compliance and lumbar compliance during lumbar infusion. This diagram illustrates the intracranial and lumbar pressure-volume curves and the relationship to the pulsatility parameters. Under normal physiologic conditions with high intracranial compliance, ICP wave amplitude is correspondingly small. As intracranial compliance decreases (steep part of the pressure-volume curve), the brain behaves increasingly like a linear elastance and so variations in intracranial volume correlate increasingly well with changes in ICP wave amplitude, thus the steepness of the pressure-volume curve accounts for large-amplitude ICP waveforms. The cranial pressure-volume curve is left of the lumbar pressure-volume curve, illustrating that compliance is higher in the lumbar compartment.