Morphological characterization of cardiac induced intracranial pressure (ICP) waves in patients with overdrainage of cerebrospinal fluid and negative ICP

https://doi.org/10.1016/j.medengphy.2011.11.011Get rights and content

Abstract

Symptomatic overdrainage of cerebrospinal fluid (CSF) can be seen in shunted hydrocephalus patients and in non-shunted patients with spontaneous intracranial hypotension (SIH). In these patients, intracranial pressure (ICP) monitoring often reveals negative static ICP, while it is less understood how the pulsatile ICP (cardiac induced ICP waves) is affected. This latter aspect is addressed in the present study. A set of 40 ICP recordings from paediatric and adult hydrocephalus patients were randomly selected. Each cardiac induced ICP wave was automatically identified and manually verified by the beginning and ending diastolic minimum pressures and the systolic maximum pressure. The ICP wave parameters (static pressure, amplitude, rise time, rise time coefficient, downward coefficient, wave duration, and area-under-curve) were then automatically computed. The material of 40 ICP recordings provided a total of 3,192,166 cardiac induced ICP waves (1,292,522 in paediatric patients and 1,899,644 in adult patients). No apparent changes in ICP wave parameters were seen when mean ICP became negative, except that the parameters amplitude, rise time coefficient, downward coefficient and area under curve somewhat increased when mean ICP was below −15 mmHg.

Introduction

The main components of the intracranial volume are the brain tissue (85%), cerebrospinal fluid (CSF; 10%) and blood (5%). The CSF has a vital role in the physical protection of the brain as well as being crucial for brain metabolism [1]. Normally the production and absorption of CSF is balanced; disruption of this balance may cause hydrocephalus with excess of CSF within the brain ventricles [2]. Hydrocephalus is often treated with surgical implantation of a shunt for drainage of CSF from the cerebral ventricles to the abdominal cavity or the heart. Overdrainage of CSF is a frequent complication to shunt treatment [3], [4], [5], as well as a complication to other types of neurosurgery and even lumbar puncture (LP) [6], [7]. When the cause of CSF overdrainage is unknown, it is denoted spontaneous intracranial hypotension (SIH) [6], [7]. Independent of the cause, the symptoms can be severe, including headache, unsteadiness, vertigo, lethargy, and cranial nerve dysfunction.

Cerebrospinal fluid overdrainage often represents a major challenge to the physician. It may cause severe morbidity to the patient, and be hard to diagnose [7]. It is usually difficult and often impossible to differentiate CSF overdrainage from underdrainage solely based on symptoms and imaging the cerebral ventricles [8]. In order to diagnose intracranial hypotension (low intracranial pressure – ICP), usually the fluid pressure is measured through a LP [6], [7]. Some previous reports indicate a useful role of continuous ICP monitoring in shunted hydrocephalus patients with possible over- or underdrainage [9], [10], [11], [12]. We found continuous ICP monitoring to be useful in shunted patients with severe and lasting symptoms not responsive to management [8].

We have observed that the ICP usually becomes negative during symptomatic CSF overdrainage [8]. However, one obstacle when referring to the ICP is that a negative ICP may also be caused by erroneous baseline pressure [13], [14]. For this reason we have incorporated assessment of the cardiac induced ICP waves since they are independent of the baseline pressure [13]. If the morphology of ICP waves relates to the degree of CSF overdrainage, it might be possible to improve diagnostics.

The aim of the present study was to characterize the morphology of cardiac induced ICP waves in patients with CSF overdrainage, either caused by shunt treatment or SIH. As reference, we also characterized the morphology of cardiac induced ICP waves in other hydrocephalus patients, both children and adults.

Section snippets

Patient material

The material consists of 40 patient recordings (11 paediatric patients and 29 adults) that were randomly selected from the Pressure Database of Department of Neurosurgery, Oslo University Hospital – Rikshospitalet. The ICP recordings had been done during the period 2002–2007, as part of diagnostic work-up.

This study was approved by the hospital authority of Oslo University Hospital – Rikshospitalet (10/16550). The Regional Committee for Research Ethics was informed about the study in writing,

Patient material

Demographic data of the patient material are summarized in Table 1. A total of 40 patient recordings were included in the study; the recordings of these patients incorporated a total of 3,192,166 manually identified cardiac induced ICP waves (1,292,522 in paediatric patients, and 1,899,644 in adult patients). Among the 40 patients the median number of single waves was 71,103; the median number of single waves among patients in each group is presented in Table 1.

ICP wave morphology – paediatric and adult patient categories

Table 2 summarizes the ICP wave

Discussion

In this observational study we describe how ICP wave morphology is changed during CSF overdrainage when ICP becomes negative. The ICP wave parameters in patients with shunt overdrainage compared best with the ICP wave morphology seen in patients with spontaneous overdrainage (or SIH), and with the preoperative ICP wave morphology in shunt non-responders. During CSF overdrainage when ICP is in the range 0 to −15 mmHg, no apparent change in ICP wave parameters was seen, while in adult the ICP wave

Conflict of interest statement

PKE has financial interest in the software company (dPCom AS) that manufactures the software (Sensometrics® Research Software and Sensometrics® Software) in this study. MS, AW, and TES report no conflicts of interest.

Acknowledgements

The authors are grateful to dPCom AS, Oslo, Norway, for providing us with Sensometrics Technology, Advanced Research Edition, and to Trond Stadheim, dPCom AS, for preparing data files.

References (17)

There are more references available in the full text version of this article.
View full text