Extrapolation to zero-flow pressure in cerebral arteries to estimate intracranial pressure

W. Buhre*,1, F. R. Heinzel2, S. Grund3, H. Sonntag3 and A. Weyland4

1 Klinik für Anästhesiologie, Universitätsklinikum der RWTH Aachen, Pauwelsstrasse 30, D-52074 Aachen, Germany. 2 Institut für Pathophysiologie, Universitätsklinikum Essen, Germany. 3 Zentrum Anaesthesiologie, Rettungs- und Intensivmedizin, Georg-August-Universität, Göttingen, Germany. 4 Klinik für Anästhesiologie und Operative Intensivmedizin Städtische Kliniken, Oldenburg, Germany

Corresponding author. E-mail: wbuhre@ukaachen.de

Accepted for publication: November 4, 2002


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background. Cerebral perfusion pressure (CPP) is commonly calculated from the difference between arterial blood pressure (AP) and intracranial pressure (ICP). ICP can be considered the effective downstream pressure of the cerebral circulation. Consequently, cerebral circulatory arrest would occur when AP equals ICP. Estimation of AP for zero-flow pressure (ZFP) may thus allow estimation of ICP. We estimated ZFP from cerebral pressure–flow velocity relationships so that ICP could be measured by transcranial Doppler sonography.

Methods. We studied 20 mechanically ventilated patients with severe head injury, in whom ICP was monitored by epidural pressure transducers. AP was measured with a radial artery cannula. Blood flow velocity in the middle cerebral artery (VMCA) ipsilateral to the site of ICP measurement was measured with a 2 MHz transcranial Doppler probe. All data were recorded by a microcomputer from analogue–digital converters. ZFP was extrapolated by regression analysis of AP–VMCA plots and compared with simultaneous measurements of ICP.

Results. ZFP estimated from AP–VMCA plots was linearly related to ICP over a wide range of values (r=0.93). There was no systematic difference between ZFP and ICP. Limit of agreement (2 SD) was 15.2 mm Hg. Short-term variations in ICP were closely followed by changes in ZFP.

Conclusion. Extrapolation of cerebral ZFP from instantaneous AP–VMCA relationships enables detection of severely elevated ICP and may be a useful and less invasive method for CPP monitoring than other methods.

Br J Anaesth 2003; 90: 291–5

Keywords: brain, blood flow; brain, intracranial pressure; head, injury


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Measurement of intracranial pressure (ICP) is important in the management of patients with severe head injury, cerebral ischaemia and subarachnoid haemorrhage.1 2 Thus, invasive measurement of ICP by epidural, intraventricular or intracerebral pressure transducers is used more often, particularly immediately after injury.13 Experi mental and clinical studies have shown the value of estimating cerebral perfusion pressure, which is the difference between the upstream pressure (mean arterial pressure, MAP) and the effective downstream pressure (EDP) of the cerebral circulation.4 In clinical practice, ICP is commonly used to represent EDP.4 This assumption is based on the fact that some cerebral veins are highly compressible and cerebrospinal fluid pressure therefore determines the postcapillary venous outflow pressure of the cerebral circulation as long as ICP exceeds central venous pressure.5 This implies that EDP is primarily determined by collapsible cerebral veins with Starling resistor properties.6 The cerebral circulation will stop if the MAP equals the EDP. We have shown that the EDP of the cerebral circulation can be reasonably assessed from instantaneous pressure–flow velocity plots by extrapolation to zero-flow pressure (ZFP).4 In contrast to invasive measurement of ICP, intracranial sensor placement is not needed and the method could be available immediately in emergencies.4 We showed that when cerebral blood flow is changed by changes in arterial carbon dioxide tension, ZFP is not the same as ICP if intracranial hypertension is absent, suggesting that a second Starling resistor is acting, probably at the precapillary, arteriolar position in the cerebral circulation.4 This precapillary Starling resistor determines EDP as long as ICP does not exceed the critical closing pressure of the arteriolar system. The concept of two Starling resistors in series predicts that the higher downstream pressure (pre- or postcapillary) determines the EDP of the cerebral circulation. Therefore, ZFP as a measure of EDP should be the same as ICP if intracranial hypertension is present. We conducted an observational study to support this hypothesis.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
After approval by the institutional review board, we studied 20 patients suffering from severe head injury. Routine ICP monitoring was done with epidural transducers. Intracranial pressure on admission was >20 mm Hg in all patients during normocapnia. During the study period, patients received analgesia and sedation using a combination of midazolam, piritramid and ketamine. Mechanical ventilation was adjusted to obtain normocapnia. Patients with ultrasound evidence of cerebral vasospasm, haemodynamic instability or severe respiratory failure (FIO2 >0.4) were excluded. The mean age was 41 (range 19–71) yr, body weight 80 (SD 9) kg and height 175 (7) cm. All but six patients received cardiovascular support with dobutamine (2–8.5  µg kg–1 min–1) and norepinephrine (0.04–0.5  µg kg–1 min–1) to provide a cerebral perfusion pressure >70 mm Hg whenever possible. Measurements were performed immediately after initial haemodynamic stabilization and repeated during the next 48 h. Arterial blood pressure, heart rate, central venous pressure and ICP were recorded continuously. The arterial blood pressure was measured from a radial artery and pressure transducers were referenced at the level of the base of the skull.

Arterial oxygen saturation and carbon dioxide tension were measured at the beginning and end of each set of measurements. Blood flow velocity was measured in the middle cerebral artery (VMCA) on the same side as the site of ICP measurement, with a 2 MHz Doppler device (TC 2000; EME, Ueberlingen, Germany). After finding the blood flow velocity curve in the middle cerebral artery, the depth of insonation was adjusted to obtain signals from the proximal segment (M1) of the middle cerebral artery.79 Then, the Doppler probe was fixed with a probe holder device (IMP monitoring probe holder; EME), and the position of the probe was left unchanged until the end of the study period. Arterial blood pressure and VMCA curves were sampled with an analogue– digital converter at a frequency of 50 Hz and the data were stored on a microcomputer for further calculation.

EDP was calculated by extrapolation of Doppler-derived pressure–flow velocity plots as described recently.4 Since cerebral blood flow stops if MAP equals EDP, EDP would be the same as arterial pressure when cerebral blood flow was zero. In patients with spontaneous circulation, ZFP cannot be measured directly; therefore ZFP was determined in the present study using instantaneous arterial pressure– flow velocity plots (AP–VMCA).4 Arterial pressure and VMCA data of single heart-beats were plotted against each other and extrapolated to zero flow using linear regression analysis.4 As pressure and flow were measured at different locations, compensation for time delay was done by iterative regression analysis until hysteresis of AP–VMCA plots was minimized.4 The pressure axis intercept of these plots represents ZFP data of the cerebral circulation.4 For each measurement, two respiratory cycles were selected randomly and extrapolated ZFP data of these heart-beats were averaged for further analysis.

Statistics
Haemodynamic data in Table 1 are given as mean (SD). Directly measured ICP and calculated ZFP data were compared according to the method of Bland and Altman.10 Bias between methods was determined as the mean difference between ZFP and ICP. The precision of both methods was given by the limits of agreement (2 SD of the difference between methods). In addition, correlation analysis was performed to assess the relationship between ICP and ZFP.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Cerebral EDP was estimated by linear regression of instantaneous pressure–flow velocity plots in a total of 180 measurements. The median number of measurements per patient was 7 (range 5 – 20). Haemodynamic values and blood gas analysis are shown in Table 1. Mean ICP and ZFP during the study period were 35.2 (18.1) and 34.7 (20.1) mm Hg respectively, indicating no significant difference between methods. The limits of agreement (2 SD) according to Bland and Altman were 15.2 mm Hg (Fig. 1). Individual differences between methods did not depend on the absolute values of ZFP and ICP (Fig. 1).


View this table:
[in this window]
[in a new window]
 
Table 1 Haemodynamic variables and arterial carbon dioxide tension
 


View larger version (18K):
[in this window]
[in a new window]
 
Fig 1 Comparison between ICP and ZFP. Mean difference between methods was 0.49 mm Hg (continuous line). Limits of agreement (2 SD) were 15.2 mm Hg (dashed lines). n=180.

 
In three patients with intracranial hypertension that did not respond to treatment, cerebral blood flow ceased during diastole. In this situation, ZFP did not have to be extrapolated from arterial pressure–VMCA plots; instead the ‘true’ ZFP could be measured from the arterial pressure curve. ZFP in these patients was similar to direct ICP measurements (Fig. 2).



View larger version (13K):
[in this window]
[in a new window]
 
Fig 2 Direct assessment of cerebral ZFP. In a patient with severe intracranial hypertension, no flow was observed in the middle cerebral artery during diastole. The epidurally measured ICP was 48 mm Hg.

 
As ZFP is measured from data of single heart-beats, the recordings allowed comparison of short-term changes in ZFP and ICP, e.g. during variations in ICP induced by the respiratory cycle. Figure 3 shows the relationship between ZFP and ICP during mechanical ventilation; the time courses are similar and there is close agreement between the two variables.



View larger version (16K):
[in this window]
[in a new window]
 
Fig 3 Variation in ZFP and ICP during mechanical ventilation. ZFP and ICP show similar temporal resolution and are influenced identically by changes in intrathoracic pressure.

 

    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
We found that (i) cerebral ZFP can be estimated by extrapolation from instantaneous pressure–flow velocity relationships in patients with severe cerebral injury, and that (ii) cerebral ZFP and ICP are closely related to each other, indicating the EDP of the cerebral circulation during intracranial hypertension.

Basic physiology predicts that blood flow through an organ will stop if the difference between the upstream and the downstream pressure is abolished; thus, the arterial pressure at zero flow will be the same as the EDP of the vascular bed. Because of the peculiar anatomy of cerebral veins and the rigidity of the cranium, the EDP of the cerebral circulation is determined by ICP if intracranial hypertension is present and indicates ZFP under these conditions.

To assess ZFP, several experimental studies have used stop-flow methods. Other studies extrapolated to ZFP by curve-fitting of pressure–flow plots. In most of these studies, paired pressure–flow data were derived from variation in blood flow by different interventions using repeated flow measurements. In our study, ZFP was measured by extrapolation from instantaneous pressure–flow velocity relationships, using the physiological variations in blood pressure and blood flow during the cardiac cycle. This method has been used previously to calculate the EDP in the coronary circulation11 12 and has been used recently to assess the EDP of the cerebral circulation in a clinical setting.1315

We found close agreement between mean ICP and mean ZFP. The lack of a systematic difference confirms the hypothesis that ZFP, as extrapolated from instantaneous pressure–flow velocity plots, represents the EDP of the cerebral circulation in patients with intracranial hypertension. The relationship between individual ICP and ZFP data additionally shows satisfactory agreement between the two methods of measurement. There may be individual differences between ZFP and ICP for different reasons.

First, extrapolation of ZFP from pressure–flow velocity plots requires high-quality recordings of both arterial pressure and middle cerebral artery blood flow velocity. In some patients poor resolution of Doppler flow measurements may cause errors in extrapolating ZFP. Damping of arterial pressure measurements may cause similar problems. As the arterial pressure curve was from the radial artery, there was a time delay between pressure and flow, and such a delay requires correction. If this correction is inaccurate there may be residual hysteresis of the pressure–flow velocity plots.4

Secondly, direct measurements of ICP may also be subject to methodological inaccuracy, which could cause differences between ZFP and ICP. In particular, epidural measurements are not accepted as a gold standard in determining ICP. A parallel study compared the EDP from pressure–flow velocity plots with ICP measured by intraventricular pressure recordings in patients with and without intracranial hypertension.16 The limits of agreement were comparable (16.2 vs 15.2 mm Hg in our patients), suggesting that individual differences between EDP and ICP are a physiological rather than a methodological problem.16 From the physiological point of view, ZFP is a measure of cerebral EDP, whereas ICP represents an indirect estimate of tissue pressure. This is true whether a pre- or a postcapillary Starling resistor determines the EDP. Particularly in patients with high cerebral vasomotor tone, EDP may be greater than ICP if critical closing pressure is located at the precapillary (arteriolar) level, as we found in patients with intracranial normotension during deliberate changes in arterial carbon dioxide tension.4 13

Czosnyka and colleagues,17 in a similar clinical study, investigated the relationship between cerebral ZFP (as a measure of critical closing pressure) and ICP. In contrast to our results, the correlation between critical closing pressure and ICP was weak (r=0.41) and the authors concluded that critical closing pressure cannot be used to estimate ICP. This difference in results may be explained by the method of ZFP determination. Czosnyka and colleagues17 calculated the zero flow intercept point by extrapolation of systolic and diastolic pressure–flow velocity data pairs only. Recording of systolic data at a sampling rate of 50 Hz may give rise to considerable errors in the extrapolation of ZFP if the remaining data from pressure–flow velocity changes during the cardiac cycle are neglected. Hysteresis of pressure–flow velocity plots because of a time delay in one of signals cannot be detected.

However, in the study of Czosnyka and colleagues17 the highest deviation between critical closing pressure and ICP was found in patients with intact autoregulation, i.e. in the presence of significant vasomotor tone; no significant difference was observed in patients with disturbed autoregulation. Correspondingly, the relative differences between ZFP and ICP in our study were greatest at low and moderate ICP values, when vasomotor tone was more likely to have a greater effect than the influence of ICP on EDP.

Extrapolation from instantaneous pressure–flow velocity relationships enables determination of EDP on a beat- to-beat basis. This gives high temporal resolution, shown by on-line recordings of short-term variations in EDP. The example in Figure 3 shows variations in ICP caused by intermittent positive pressure ventilation closely paralleled by changes in ZFP. The temporal resolution of ZFP from instantaneous pressure–flow relationships is sufficient to detect cyclic changes in EDP, e.g. abrupt increases in ICP (A-waves), which are of pathophysiological relevance.

In a number of patients with refractory intracranial hypertension, zero flow in the cerebral circulation occurred during diastole. The recording of pressure–flow velocity plots in these patients enabled validation of our technique of curve-fitting and extrapolation as the ‘true’ ZFP data, i.e. the arterial pressure that co-indicates complete cessation of blood flow in the middle cerebral artery. The analysis of these plots shows that the linearity of the arterial pressure–VMCA relationship is maintained even at very low flow velocities. The close agreement between the ‘true’ ZFP and ICP adds further evidence that ICP at extreme levels of intracranial hypertension indeed represents the EDP of the cerebral circulation.

The clinical values of ZFP extrapolation are several. First, transcranial Doppler sonography and arterial pressure monitoring are only moderately invasive, so that ICP can be estimated without breaching the cranium. This may be particularly useful if ICP measurement is not readily available or contraindicated because of coagulation disorders. Furthermore, more than one-third of patients with severe head injury develop intracranial hypertension within the first 24 h without signs of increased ICP in the initial CT scan.2 18 Patients with cerebral injury could be screened by quantitative assessment of EDP to examine the indication for direct ICP measurement.

In summary, we conclude that ZFP and ICP do not differ systematically and are closely related to each other in patients with severe head injury associated with moderate to severe intracranial hypertension. However, in the subgroup of patients with moderately elevated ICP, ZFP and ICP may be different entities. Estimation of ZFP from instantaneous pressure–flow velocity plots allows clinical detection of severe intracranial hypertension if other methods fail or are not available. From a physiological point of view, ZFP may, in principle, be the more rational measure to assess cerebral downstream pressure and estimate cerebral perfusion pressure.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1 Lang EW, Chesnut RM. Intracranial pressure and cerebral perfusion pressure in severe head injury. New Horiz 1995; 3: 400–9[Medline]

2 Juul N, Morris GF, Marshall SB, Marshall LF. Intracranial hypertension and cerebral perfusion pressure: influence on neurological deterioration and outcome in severe head injury. The Executive Committee of the International Selfotel Trial. J Neurosurg 2000; 92: 1–6

3 Lannoo E, Colardyn F, De Deyne C, Vandekerckhove T, Jannes C, De Soete G. Cerebral perfusion pressure and intracranial pressure in relation to neuropsychological outcome. Intensive Care Med 1998; 24: 236–41[CrossRef][ISI][Medline]

4 Weyland A, Buhre W, Grund S, et al. Cerebrovascular tone rather than intracranial pressure determines the effective downstream pressure of the cerebral circulation in the absence of intracranial hypertension. J Neurosurg Anesthesiol 2000; 12: 210–6[CrossRef][ISI][Medline]

5 Dewey RC, Hunt WE. Cerebral hemodynamic crisis. Physiology, pathophysiology, and approach to therapy. Am J Surg 1976; 131: 338–49[CrossRef][ISI][Medline]

6 Melot C, Delcroix M, Closset J, et al. Starling resistor vs. distensible vessel models for embolic pulmonary hypertension. Am J Physiol 1995; 268: H817–27[Abstract/Free Full Text]

7 Weyland A, Stephan H, Kazmaier S, et al. Flow velocity measurements as an index of cerebral blood flow. Validity of transcranial Doppler sonographic monitoring during cardiac surgery. Anesthesiology 1994; 81: 1401–10[ISI][Medline]

8 Aaslid R, Markwalder TM, Nornes H. Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg 1982; 57: 769–74[ISI][Medline]

9 Aaslid R, Huber P, Nornes H. A transcranial Doppler method in the evaluation of cerebrovascular spasm. Neuroradiology 1986; 28: 11–6[ISI][Medline]

10 Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–10[ISI][Medline]

11 Nanto S, Masuyama T, Hori M, Shimonagata T, Ohara T, Kubori S. Zero flow pressure in human coronary circulation. Angiology 1996; 47: 115–22[ISI][Medline]

12 Klocke FJ, Mates RE, Canty JM Jr, Ellis AK. Coronary pressure–flow relationships. Controversial issues and probable implications. Circ Res 1985; 56: 310–23[Abstract]

13 Early CB, Dewey RC, Pieper HP, Hunt WE. Dynamic pressure-flow relationships of brain blood flow in the monkey. J Neurosurg 1974; 41: 590–6[ISI][Medline]

14 Dewey RC, Pieper HP, Hunt WE. Experimental cerebral hemodynamics. Vasomotor tone, critical closing pressure, and vascular bed resistance. J Neurosurg 1974; 41: 597–606[ISI][Medline]

15 Panerai RB, Kelsall AW, Rennie JM, Evans DH. Estimation of critical closing pressure in the cerebral circulation of newborns. Neuropediatrics 1995; 26: 168–73[ISI][Medline]

16 Thees C, Scholz M, Schaller MDC, et al. Relationship between intracranial pressure and critical closing pressure in patients with neurotrauma. Anesthesiology 2002; 96: 595–9[ISI][Medline]

17 Czosnyka M, Smielewski P, Piechnik S, et al. Critical closing pressure in cerebrovascular circulation. J Neurol Neurosurg Psychiatry 1999; 66: 606–11[Abstract/Free Full Text]

18 Stocchetti N, Penny KI, Dearden M, et al. Intensive care management of head-injured patients in Europe: a survey from the European brain injury consortium. Intensive Care Med 2001; 27: 400–6[CrossRef][ISI][Medline]