Effect ofsupratentorial space-occupying lesions regional ...Effect ofsupratentorial space-occupying lesions on regional intracranial pressure andlocal cerebral blood flow: anexperimental

Journal of Neurology, Neurosurgery, and Psychiatry, 1974, 37, 617-626

Effect of supratentorial space-occupying lesions onregional intracranial pressure and local cerebralblood flow: an experimental study in baboons'

L. SYMON2, E. PASZTOR3, N. M. BRANSTON, AND N. W. C. DORSCH

From the Departmient of Neurosurgical Studies, The Inistitute of Neurology,Queen Square, Loncdon

S Y N OPSIS Cortical blood flow and epidural intracranial pressure have been measured in the twosupratentorial compartments of the intracranial space in experimental baboons during the acuteexpansion of a parieto-occipital epidural balloon. Differential pressures between the two halves ofthe supratentorial space have been found, and these have been associated with evidence that flow hasfallen more quickly in the hemisphere most compressed. The evidence points to a more rapidexhaustion of the autoregulatory capacity in the hemisphere subjected to greater compression, a fallin perfusion pressure to below critical autoregulatory levels occurring slightly before that in theopposite hemisphere, and the establishment of a differential flow pattern for a short time during a

critical phase of compression. The displacements induced by inflation of the parieto-occipital balloonhave been described.

The relationship between the failure of functionof nervous tissue and the development of a com-pressing lesion is of primary interest to the neuro-surgeon. Investigation of the pathophysiologicalprocesses involved in this relationship has beenstimulated increasingly in recent years by de-velopment of techniques for the measurement ofcerebral blood flow in a regional (Lassen et al.,1963; H0edt-Rasmussen et al., 1966; Hoedt-Rasmussen, 1967) or a focal manner (Auklandet al., 1964; Aukland, 1965), and by the accumu-lation of data on intracranial pressure obtainedeither in the clinic using a central cerebrospinalfluid (CSF) space measurement (Lundberg, 1960)or using transducers inserted into the subdural(Hulme and Cooper, 1966; Tindall et al., 1972)or epidural space (Jacobson and Rothballer,1967; Dorsch et al., 1971; Sundbarg andNornes, 1972; Symon et al., 1972). It has be-

The study was supported by the Medical Research Council and bythe Wellcome Foundation.2 Address for reprints: Mr. Lindsay Symon, FRCS, Department ofNeurosurgical Studies, The National Hospital, Queen Square, LondonWCIN 3BG, England.3 Dr. Pasztor was the recipient of a research fellowship from theWellcome Foundation. Present address: The Institute of Neuro-surgery, 1145 Budapest, Amerikai -uS1, Hungary.

617

come clear that the type of neurological dys-function resulting from a space-occupying lesionis critically dependent not only on the site of thelesion but on the rapidity of its growth. To someextent this is a reflection of the rate at which thelesion crosses the flat or accommodating part ofthe pressure/volume curve (Langfitt, 1969), butit is only by the accumulation of data acquiredby modern methods that a meaningful interpreta-tion of the relationships between circulatorydepression secondary to pressure, neurologicaldysfunction secondary to circulatory impair-ment, and the origin of tissue shifts within thebrain can begin.The present work was an attempt to relate

pressure and flow events occurring in response torapidly inflated epidural expanding lesions inprimates.

METHODS

Six baboons (Papio cynocephalus or Papio nubius) ofeither sex, in the weight range 10-15 kg, were used inthese experiments. After tranquillization withphencyclidine, they were anaesthetized with a sleepdose of thiopentone, followed by intravenous a

Protected by copyright.

on May 20, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.37.6.617 on 1 June 1974. Dow

nloaded from

http://jnnp.bmj.com/

L. Symon, E. Pasztor, N. M. Branston, and N. W. C. Dorsch

chloralose (60 mg/kg), intubated, and respirationmaintained with a Starling pump after immobiliza-tion with gallamine triethiodide (1 mg/kg repeatedas necessary). End-tidal CO2 was continuouslymonitored with a Beckman gas analyser (modelLB1) by sampling from the endotracheal tube. Bloodpressure was recorded by a P23G Statham arterialgauge from a femoro-aortic catheter, central venouspressure recorded by a P23V Statham gauge fromthe superior vena cava via one femoral vein, and asecond femoro-aortic catheter introduced for inter-mittent sampling of arterial blood for analysis in themicro-Astrup apparatus for pH, pCO2, and PO2.The respirator volume was adjusted to maintain anormal arterial pCO2 between 35 and 45 mmHg, andpure oxygen used as the inhalational gas. Theanimals were then positioned in a Narishige stereo-tactic apparatus (model SN3) and twist drill holes inthe skull made to enable placement of fine platinumelectrodes (Symon et al., 1973) in putamen andwhite matter on both sides, and in frontal andparietal cortex bilaterally. The reference electrodefor the hydrogen system was placed subcutaneouslyin the interscapular region. Four extradural pressuregauges (Dorsch et al., 1971) were placed through12 mm trephine holes in both frontal and parietalzones, and a further 12 mm trephine hole made inthe postparieto-occipital zone on the right side. Thetransducers were held in place with acrylic, and anextradural balloon was introduced to lie in theparieto-occipital zone near the occipital pole of theright hemisphere through the parieto-occipitaltrephine hole, the dura mater having been stripped.This defect was then closed with acrylic.The transducers used were similar to those pre-

viously used (Come et al., 1968) with two majormodifications. The construction of the presentgauges was entirely coplanar, the tip of the sensingarm being on the same plane as the outer and sensi-tive part of the body, thus eliminating any influenceof dural tension (Dorsch and Symon, 1972; Majorset al., 1972). In these gauges also, metal foil ratherthan semiconductor gauges were used, so that fullbridge configuration within the gauge could beemployed, with resulting improvement in sensitivityand drift. A spinal needle was introduced into thecisterna magna and connected with a StathamP23 BB pressure transducer for recording of CSFpressure. After a few minutes had been allowed forpressures to stabilize, the extradural pressures werecalibrated against cistemal pressure as an extracheck, although extradural pressure could be ob-tained by the substitution of current deflection in apreviously made in vitro calibration graph. Extra-dural pressure calibrated in this way correlated wellwith cistemal pressure measured at the same time

mm Hg35

30

25

20

15

10

5

0

y

y-098x.O-24, / r- 0-9994

/ SEx- 0 297l/ SEy 0292

///

/ x0 5 10 15 20 25 30 35 mm Hg

FIG. 1. A comparison of cisternal (x) and extradural(y) pressures in a baboon. Extraduralpressures in thisgraph are predicted from the preliminary in vitrocalibration of the transducer, and the 95Y0 confidencelimits are shown.

(Fig. 1), providing that accurate coplanar placementof the extradural transducer was ensured andexcessive dural indentation avoided. As a check,therefore, intracranial pressure was in each caseraised by between 30 and 50 mmHg by the inhalationof C02, and deflection of extradural recordingsplotted against simultaneous cisternal pressuremeasurements during both the rising and fallingphases of intracranial pressure. A graph of deflectionof extradural pressure against intracranial pressurewas then made for use during the latter part of theexperiment (Fig. 2). This CO2 calibration was com-bined with control and test episodes of hydrogeninhalation, 7%.-I0%O hydrogen being introducedinto the endotracheal tube for three to four minutesper saturation (Pasztor et al., 1973). Hydrogen clear-ance was measured for the 10 minutes after the dis-continuance of each hydrogen inhalation. Cisternalrecording was maintained for as long as possible,but invariably failed in the course of the experimentwith the prolapse of the cerebellar tonsils usually atan intracranial pressure exceeding 70 mmHg.The balloon inflated at a constant rate of 0-2 ml/

min by a Palmer slow infusion pump, which rancontinuously from the start of inflation of theballoon except during episodes of hydrogen clear-ance, when, in order to avoid irregularities in theintracranial pressure and to minimize systemic bloodpressure reactions, the inflation of the balloon was

618



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Regional intracranial pressure and local cerebral bloodflow

casternd Pressure mm Hg

FIG. 2. In vivo calibration ofthe output ofthe extra-dural transducer (expressed as millimetres of pen

deflection on a Beckman type R recorder) plottedagainst cisternal pressure in mmHg. A high degree ofcorrelation is evident (P< 0001).

maintained intermittently to keep the raised level ofintracranial pressure at a constant level duringhydrogen clearance. Escape of fluid from the balloonwas not possible in the course of the experiment.The outputs of the intracranial pressure trans-

ducers and of the systemic blood pressure and CVPmonitors were displayed on a Beckman type R poly-graph, the outputs of the hydrogen electrodes being

recorded on a Rikadenki 4-channel slow recorder(model B4021), each recorder channel sampling twoelectrodes alternately. The pupillary reactions wererecorded by direct examination in the course of aballoon inflation. When inflation of the balloon hadbeen pursued to high levels and very slow clearancewas being obtained from both hemispheres, theexperiment was terminated in one of two ways. Inthe first, hyperaemia was induced by suddenly de-flating the balloon, and the effect recorded by afurther episode of hydrogen clearance (Symon et al.,1973). The balloon was then reinflated with exactlythe same volume and the animal killed. In the second,the balloon remained inflated and the animal waskilled. Finally, in all experiments, the head of theanimal was removed and fixed with the extraduralballoon in situ after removal of transducers andelectrodes, so that shifts induced by the extradurallesion and mass depression of the occipitoparietalcortex could be examined at leisure.

INTRACRANIAL PRESSURE MEASUREMENTS The datafor intracranial pressure measurements, taken duringthe course of gradually increasing pressure inducedby inflation of the extradural balloon, are shown inTable 1 for the four transducers in the six experi-ments. For the purpose of analysis, four pointsduring the evolution of the experiment have beenchosen. The first of these was when the cisternalreference external transducer recorded a pressure of10 mmHg. This was usually slightly into the experi-ment from baseline but constituted a handy point of

TABLE 1INTRACRANIAL PRESSURE DATA*

ICP level Experiments Mean SD Hemisphere P

1 2 3 4 5 6 Mean SD

Cisternal L anterior 9 0 10 0 5-0 8-5 6-0 8-0 7-75 1-89pressure L posterior 9-5 13-0 6-5 10 0 9 0 6-0 9 00 2 55 8-83 2-1410 mmHg R anterior 6 5 9 5 6-0 10 0 9 0 8-0 8-17 1-63 l 10 30 J > 0-05

R posterior 16-0 11 0 5 0 12-0 19.0 11 5 12 42 4-78 J

Left posterior L anterior 39 5 40 0 29 5 27-5 28-6 27-0 32-02 6 05extradural L posterior 30-0 300 30-0 30 0 300 30 0 3000 0o00 } 31-01 4-04pressure R anterior 37 5 48-0 28-0 39-0 30-0 32 5 35-83 7-31 <0-0130 mmHg R posterior 52 0 55-0 37 5 43 0 37-6 33 0 43-02 8-77 39J 4 820

Left posterior L anterior 65 0 79 7 61-5 75 0 57 0 70 0 69-03 7-92extradural L posterior 70 0 70 0 700 70-0 70 0 70 0 70 00 000 5 69-52 5-14pressure R anterior 91.0 81-0 71-0 78-6 58 0 92-0 78 60 12-82 8349 1067 000170 mmHg R posterior 91.0 96-0 77-5 88-0 82-7 95 0 88-37 7-20 C1

Peak level L anterior 83-0 132-0 1900 80-0 80-0 153-0 119-67 46-26left posterior L posterior 86-0 113-0 189-0 84-0 104-0 149-0 120-83 40 88 1-20-25 39 85extradural R anterior 104-0 127-0 197l0 100 0 86-0 192-0 134-33 48 46 14717 4251pressure R posterior 116-0 164-0 207-0 160-0 113-0 200-0 160-00 }

* Extradural pressure in mmHg.

619



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



reference. It was thereafter unsatisfactory to use thecisternal reference, since this pressure recording in-variably failed at some stage during the developmentof the experiment. The left posterior transducer, thatis, the one furthest away from the balloon on theopposite side of the falx, was thereafter chosen as thereference. Another comparison was made when thisreference transducer recorded a level of 30 mmHg,again at 70 mmHg, and finally when the intracranialpressures had reached their peak and blood flowswere very low or absent.

Early in the development of the experiment, withthe cisternal pressures recording 10 mmHg, there wasno significant difference between the recordings fromthe four transducer sites considered individually.The right posterior transducer anterior to the area ofthe balloon insertion usually recorded slightly highervalues than the rest, the mean being 12-42 mmHgcompared with the other means of 9 mmHg for theleft posterior, 7-75 mmHg for the left anterior and8-17 for the right anterior, but these differences werenot significant. The hemispheral mean pressures atbasal level, 8-83 mmHg (SD ± 2-14) for the lefthemisphere, and 10-30 mmHg (SD ± 3 89) for theright hemisphere, were also not significantly differ-ent.By the time the left posterior or reference trans-

ducer had reached 30 mmHg, the difference betweenthe right and left-sided transducer pressures hadreached significant levels, the right-sided mean being39.43 mmHg (SD + 8 2) and the left-sided mean31 01 mmHg (SD ± 4 0). The difference between thetwo was significant (P < 0-01). As the intracranialpressure continued to rise, reaching the left posteriorreference of 70 mmHg, differences between the twosides increased. Thus, the mean left-sided extraduralpressure of 69-52 mmHg (SD ± 5 1) and the meanright-sided pressure of 83-49 mmHg (SD± 10 7)were significantly different at the P < 0 001 level. Aspressures approached peak values and clearancelevels declined, there remained significantly differentpressure levels on the two sides: mean left-sidedpressure of 12025 mmHg (SD ± 39-9) comparingwith mean right-sided pressure of 147 17 mmHg(SD ± 42 5), with P < 0 01. It is interesting that,though the right posterior transducer invariablyrecorded higher pressure than the others, the scatterof the data meant that a statistically significantgradient between this and the others could not beobtained. It did not prevent the clear demonstration,however, that the pressures in the supratentorialcompartment on the side of the balloon becamesignificantly higher than those on the opposite sideof the falx.

In one experiment (P.F.B. 3), the position of theballoon was slightly different from the other five;

what was intended as a strictly parieto-occipitalplacement on the right side developed in the courseof the experiment into a mainly right-sided place-ment, with some extension across the midline behindthe sagittal sinus. The figures for this experiment,which are included in the general table, are ofinterest in that the differences between the right-sided transducers and the left-sided transducers aresomewhat lower than in the other experiments.

EFFECTS OF EXTRADURAL EXPANDING LESION ON BLOODFLOW The effect of the gradually expanding extra-dural lesion in the right parieto-occipital zone onblood flow in the two halves of the supratentorialcompartment is summarized in Table 2. Although

TABLE 2FLOW DATA

Experiment ICP level (LPE-D transducer)(flow in ml/100 glmin)

10 mmHg 30 mmHg 70 nmmHg Peak

Left Right Left Right Left Right Left Right

1 80-6 99 0 69-3 75-3 83-4 83-4 11-0 6-92 75-3 57-7 63-0 57-7 26-6 3-4 8-7 2-33 73-8 76-0 80-6 84-9 77-0 73-8 15-4 10-04 990 73-8 86-6 69-3 86-6 57-7 71 7 4955 63-0 76-2 86-6 53 3 69-3 46-2 33 0 10 56 69-3 86-6 78-7 57-7 78-7 69-3 31*5 23-1

Mean 76-8 78-2 77-5 66-4 69-6 55-6 28-6 17-8SD 12-4 13-8 9 5 12-3 22-3 28-7 23-5 17-3P > 005 > 005 <005 <005

anterior and posterior cortical hydrogen electrodeswere placed in both hemispheres, anterior place-ments only have been used in the comparison of thecortical blood flows in the current table, since theright posterior electrode failed in a sufficiently largenumber of instances to prevent meaningful analysisof the data. Deep electrode placements will beanalysed in a later paper. Values for cortical bloodflow were values from the fast component unless amonoexponential curve was obtained, when themean blood flow gave values similar to that of thefast component (Pasztor et al., 1973). The initialvalues in the two hemispheres were similar, a meanflow of 76-8 ml/100 g/min (SD ± 12-4) for the lefthemisphere and 78-2 ml/100 g/min (SD ± 13-8) forthe right hemisphere being obtained. At an intra-cranial pressure of 30 mmHg, although the hemi-spheral blood flow on the right had fallen to 66-4ml/100 g/min (SD ± 12-3), this was neither signifi-

620



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Regional intracranial pressure and local cerebral bloodflow

cantly different from the control value nor from theleft hemispheral value of 77-5 ml/100 g/min (SD± 95). By the time the intracranial pressure hadreached 70 mmHg, blood flow in both hemisphereshad shown a slight fall, the left to 69-6 ml/100 g/min (SD ± 22 3) and the right to 55 6 ml/100 g/min(SD ± 28 7). The difference between the left hemi-spheral blood flow at 70 mmHg and control was stillnot significant, but a significant difference at the 5%/level had appeared between the right and left hemi-spheres, and the right hemisphere had now signifi-cantly lower blood flow than its own control values,again at the 500 level. At peak level of intracranialpressure, when the clearance on the right hemispherehad reached very low levels, an average mean bloodflow on the right of 17-8 ml/100 g/min (SD ± 17-3)had been reached. Blood flow in the left hemispherehad by now also declined to a mean of 28-6 ml/100 g/min (SD ± 23 5), significantly different at theP <005 level from the flow in the right. In bothinstances also, there was ahighly significant difference(P < 0-01) between the peak blood flow in the twohemispheres and control levels on each side.

EFFECTS OF EXTRADURAL SPACE OCCUPATION ON PER-

FUSION PRESSURE Perfusion pressure was calculatedas the difference between mean systemic blood

TABLE 3PERFUSION PRESSURE DATA

Experi- ICP level (LPE-D transducer)inent (perfusion pressure in mmHg)

10 ,nmHg 30 mmHg 70 mmHg Peak

Left Right Left Right Left Right Left Right

1 96 94 86 76 105 82 45 202 91 91 93 91 87 71-5 51 223 101 100 125 120 57 43 18 94 106 106 84 73 50 38 44 265 126 126 122 123 71 68 28 206 98 97 109 101 52 42 37 0

Mean 103 0 102-3 103-2 97 3 70 3 58 4 37-3 16-2SD 12 33 12-69 18 06 21-32 2194 19 68 12 25 9-72p >0-05 <0-05 <0 01 <0 01

pressure and average supratentorial compartmentpressure on each side, the mean of anterior andposterior extradural pressures being used to calculateeach average supratentorial compartment pressureat any given time (Table 3). The initial perfusionpressure was similar on the two sides of the head,being 102-3 mmHg (SD ± 12-7) on the right, and 103mmHg (SD ± 12-3) on the left. This basal level

corresponded to 10 mmHg level from the cisternalpressure transducer and represents the same point asextradural pressure records at this level in Table 1.As intracranial pressure rose to 30 mmHg, measuredfrom the left posterior transducer, perfusion pressureon the right side of the head fell to 97-3 mmHg (SD± 21 -3), while that on the left remained unchangedat 103-2 mmHg (SD ± 18-1). This was already asignificant deviation from normal on the right side(P < 0 05) and there was a significant differencebetween the two hemispheres again at the 5%/ level.By the time the intracranial pressure at the leftposterior transducer had risen to 70 mmHg, a widergap in available perfusion pressure had appeared. Onthe right, perfusion pressure had fallen to 58 4 (SD± 19-7) and that on the left had also fallen to 70 3mmHg (SD ± 21-9). Differences between the twohemispheres were now significant at P < 0-01, andboth were significantly different from the restinglevel (P < 0-01). As peak intracranial pressure wasreached, perfusion pressure levels of 16 mmHg (SD± 9 7) were obtained from the right hemisphere, and37-2 mmHg (SD ± 12 2) for the left, the differencesagain remaining significant at the P < 0-01 level.

EVIDENCE OF INTRACRANIAL SHIFTS DURING PARIETO-OCCIPITAL BALLOON INFLATION The shift induced bythe expanded balloon was checked in each instance byfixation of the head after the death of the animal, theballoon being inflated at the time. There were minordifferences in the extent of intracranial shifts pro-duced, largely dependent upon minor variations inthe siting of the extradural lesion, but in only one ofthe experiments were these differences of any signifi-cance. This was in experiment P.F.B. 3, in which theballoon had crossed the sagittal sinus and produced alesser extradural expansion in the left hemisphere. Asa result of this, the midline shift induced in thisexperiment was slightly less than in the other experi-ments, but basic characteristics were otherwise thesame.

Typical displacements are shown in Figs 3 to 5,taken from the same experiment (P.F.B. 1). Theballoon, whose volume was between 6 and 8 mlfully inflated, induced great depression in the pos-terior parietal and occipital regions centred just infront of the lunate sulcus, not crossing the midline,clearly not obstructing the sagittal sinus, and pro-ducing a conspicuous herniation of the posteriorpart of the hippocampal gyrus and the associatedtemporal lobe through the posterior part of thetentorial hiatus (Fig. 4). In some experiments, alesser contralateral herniation was also observable atthis point. As shown in Fig. 5, in addition to theforaminal hernia, appreciable herniation of theposterior part of the cingulate gyrus beneath the falx

621



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



FIG. 3 (left) P.F.B. no. 1. The parieto-occipital depression produced over one hemisphere by the inflation ofan extradural balloon (volume 8 ml).FIG. 4 (right) P.F.B. no. 1. Downward depression of the right uncus, depression of the tentorial surface of thecerebellum, and herniation of both cerebellar tonsils induced by extradural balloon inflation.

FIG. 5. P.F.B. no. 1. Coronal sections of the hemispheres, cerebellum, and brain-stem in an animal killed withmaximal inflation of an extradural balloon in the parieto-occipital zone. The brain was fixed in situ. Thereferences in millimetres are to distances posterior (negative distance) and anterior (positive distances) to theexternal auditory meatus, the point offixation in the stereotaxic apparatus.

622



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Regional intracranial pressure and local cerebral blood flow

was evident, and the posterior part of the thalamus,diencephalon and upper brain-stem was conspicu-ously deviated across the midline. The whole ven-tricular system was flattened and displaced to theleft. The cerebellar tonsils were bilaterally herniateddownwards as shown in Fig. 4, and there was noevidence of particular laterality about the posteriorfossa displacements. It is interesting that in all theanimals, appreciable down-curving of the tentoriumhad taken place, with the development of a notablyconcave upper surface to the cerebellum, the vermisbeing retained as a sharp peak.The distortions in Figs 3-5 represent the situation

in the hemisphere at the time of peak recording ofintracranial pressure and blood flow, when quitedemonstrable and significant differences in thesupratentorial compartment pressures were evident,when perfusion pressure in the two hemispheres wassignificantly different, and when there was signifi-cantly lower blood flow in the right hemisphere thanin the left.

DISCUSSION

It is only in recent years with the development ofthe Xenon clearance technique (Lassen et al.,1963; Hoedt-Rasmussen et al., 1966; Hoedt-Rasmussen, 1967) and its use in man that thewide variations in regional cerebral blood flowfrom one area of brain to the next have beenappreciated. A study of regional changes in themuch smaller primate brain has been less easy toobtain in absolute terms, since the collimation ofthe Xenon clearance technique makes multipleregional recording difficult, and since the con-tinuous recording methods of heat clearance aredifficult to quantitate. The development of thehydrogen clearance technique by Aukland andhis associates (Aukland et al., 1964; Aukland,1965), and its modification in our laboratory(Pasztor et al., 1973; Symon et al., 1973) haveenabled us to record blood flow in discrete areasof the primate hemisphere and to compare supra-tentorial cortical flow between one hemisphereand the other without any possibility of con-tamination either from extracerebral sources orfrom grey matter of the opposite side or deepgrey matter in the same hemisphere.While developments in blood flow method-

ology have reached the stage of fairly widespreadacceptance, measurement of intracranial pres-sures in the various portions of the intracranial

space has had less wide documentation anddebate continues as to how far the varioustechniques of measurements may yield com-parable results. Several groups have presenteddata (Dorsch and Symon, 1972; Majors et al.,1972; Sundbarg and Nornes, 1972; Tindall et al.,1972) to show that in conditions of low andmoderate intracranial pressure, extradural pres-sure and central CSF space pressure (cisternalor ventricular) yield identical recordings, butthat in circumstances of high intracranial pres-sure extradural recording with a coplanar devicewill yield consistently higher pressures, indicatingthat it is at this stage measuring a component ofactual tissue pressure transmitted from the brainheld under tension by the elastic pia mater itself.To discuss the potential of coplanar measure-ments for dynamic deformation analysis in thebrain would be outside the scope of this paper,but has been initiated in the work of Schettini etal. (1971), who were the first to point out theextreme importance of coplanarity in suchpressure recordings. The extradural transducermade and used in our laboratory enables record-ings to be made with ease at four points in thebaboon skull; further miniaturization of thetransducer might enable more recordings to bemade, but in the present study a four-pointplacement was sufficient to demonstrate beyondequivocation the development of differentialpressures between the supratentorial compart-ments with a posterior parieto-occipital extra-dural expanding lesion.The critical rate of lesion expansion, 02 ml per

minute, was chosen somewhat arbitrarily to pro-duce the complete evolution of critically raisedpressure within a comfortably measurable experi-mental time, but falls well within the limitsdemonstrated by Nakatani and Ommaya (1972)who showed that the development of the typicalCushing response was dependent upon the rate ofexpansion of an epidural balloon above a criticalrate which they found to be 002 ml per minutein the rhesus monkey. More gradual expansionled to prolongation of the experiment, but ifexpansion rates were cut down to below 002 ml.per minute, Nakatani and Ommaya demon-strated that differential pressures between supra-tentorial and infratentorial compartments didnot develop, and the vasopressor response couldnot be elicited. Our present work demonstrates

623



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



that in the non-compressed hemisphere there isno evidence of dissociation between anterior andposterior pressures, but (although the data arenot statistically conclusive) they do suggest thatthe differential pressure develops in the supra-tentorial compartment between the posterior andanterior portions of the compartment as thepressure increases. It seems probable, therefore,that the transmission of pressure is from theimmediate area of the balloon through thenormal hemisphere, then beneath the falx, anddiffusely into the opposite hemisphere. Our datasuggest that differential pressures are most easilyobtained with rapid rates of expansion, againagreeing with the observations of Nakatani andOmmaya (1973). The differential pressuresdemonstrated in our experiments were associatedwith shifts which would be regarded as typicalof a rapidly expanding supratentorial mass inman: subfalcine herniation, ipsilateral uncalherniation, and, to a lesser extent, contralateraluncal herniation, and bilateral foraminal impac-tion of the tonsils. The fact that these shifts werein our animals invariably associated with Cushingresponses, and with dilatation of the ipsilateralpupil initially, leads us strongly to suppose thatthe origin ofthe Cushing response is deformationof the upper brain-stem. Nakatani and Ommaya'sobservation that very slow rates of expansion ofan extradural balloon led to the disappearanceboth of the Cushing response and of differentialpressures between supratentorial and infra-tentorial compartment is, we feel, strong evi-dence in this regard also.A certain volume of work has been presented

in recent years (Johnston and Rowan, 1974)which has failed to demonstrate differential pres-sures between the supratentorial compartmentswith experimental balloon inflation, and suggest-ing that the Cushing response has its origin insome central neurogenic mechanism as yet un-determined. It seems, however, beyond doubtthat the recording of central CSF space pressuremade in such experiments is invariably that ofmean intracranial pressure, since, as long as theCSF spaces are in continuity one with the other,the pressures will be instantaneously equalizedby fluid transmission. It is possible also that theprofound collapse of the ventricular systems, asillustrated for example in Fig. 5, means that atsuch high levels of intracranial pressure satisfac-

tory pressure recording from a fluid transmissionsystem will never be practicable. In this respect,our experiences support this conclusion; thecisternal pressure invariably failed when tonsillarherniation began. It is also probable that themethodology employed in the work of Johnstonet al. (1972), with regard to blood flow analysis,is unsuitable for the determination of differentialblood flow. The use of even accurately collimatedXenon probes cannot exclude contaminationfrom the opposite hemisphere and even from thebase of the skull, and, since the differentialflows in our own series were difficult to demon-strate at the 500 level, we suggest that the addi-tional scatter introduced into the data by suchcontamination when the Xenon method wasemployed would make statistically significantdifferences very difficult to find.The effect of the expansion of the balloon

immediately beneath a Xenon collimator, as inthe Glasgow studies, must also cause concernabout changes in count rate. If, for example, aposteriorly placed collimator is trained on thepost-parietal region and the balloon is subse-quently inflated beneath this collimator to avolume of 10-14 ml saline (Johnston et al.,1973), the count rate of this detector will declinesubstantially since the tissue to which it ismainly reactive has been displaced outside thearea of maximum sensitivity of the detector.The reliability and significance of flows esti-mated using this detector will then be question-able compared with those obtained from acomparably placed contralateral probe.The finding in the present series of experiments

that blood flow did not deviate significantlyfrom normal in either hemisphere until an intra-cranial pressure of over 70 mmHg had beenreached is in keeping with our previous demon-stration that the autoregulatory characteristicsof the baboon cerebral circulation, as assessed byhydrogen clearance, will maintain normal flowto a level of intracranial pressure of this magni-tude assuming normal blood pressure in theanimal at the time. Indeed, a plot of data fromboth cortical and deep flow in the present seriesof experiments showed a curve of autoregulationsimilar to that obtained in elevation in intra-cranial pressure by cisternal infusion. In particu-lar, there was no evidence of hyperaemia eitherin superficial or deep tissues preceding the fall of

624



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Regional intracranial pressure and local cerebral blood flow

flow induced by a rising perfusion pressure(Symon et al., 1973). In the data from theexperiments as a whole it was impossible toseparate the curve of autoregulation for theearlier compressed hemisphere from that of theopposite side, but, as Tables 2 and 3 show, theappearance of significant differences in the flowdata occurred at the 70 mmHg reference level inthe left posterior transducer. At this time, thelesser compressed hemisphere retained a corticalblood flow of 69-6 ml/100 g/min, which was notsignificantly different from control, and perfu-sion pressure in this hemisphere was 70 3mmHg, within the autoregulatory range of thecirculation. On the right side, however, thehemisphere beneath the balloon had a perfusionpressure which at 58 4 mmHg is below the levelat which one would expect autoregulation toextend in this preparation, and indeed bloodflow in the cortex of this hemisphere had fallento 55-6/ml/100g/min, significantly different fromcontrol values of that hemisphere at the 500level, and from the blood flow of the oppositehemisphere at the same level of significance.From this point onwards in the experiment, bothhemispheres had perfusion pressures of a levelconsistent with a passive pressure relationship,so that the continuing difference in flow betweenthe two hemispheres is explicable by the higherintracranial pressure induced in the supratentorialcompartment on the side of the balloon, and thecorrespondingly reduced perfusion pressure inthat hemisphere as compared with its fellow. Thesupratentorial compartmental pressures at therate of balloon expansion chosen differed bylevels sufficiently great to be easily determined byan extradural pressure recording device; even so,these differences were hardly sufficient to exhaustthe autoregulatory capacity of the more com-pressed hemisphere very much before that of theless compressed. The technique of episodic bloodflow recording, therefore, may render it quitedifficult to pick up depression of flow in onehemisphere significantly before the other hemi-sphere follows suit, and it is therefore notpossible to separate curves of autoregulation forthe two hemispheres. It seems that failure oautoregulation in our preparation took place inone hemisphere shortly before the 70 mmHgintracranial pressure was reached at the referencetransducer, and in the other hemisphere shortly

after this point. Perhaps a more frequent re-cording of flow by the technique of twominute clearance might resolve this close flowmatching.

It is also difficult to resolve events at or aboutthe critical level of failure of autoregulation, inrelation to the development of intracranial shiftsand of Inedullary compression. In our ownexperiments, about the time when intracranialshifts were occurring, as deduced from failure ofreaction and beginning of dilatation of the ipsi-lateral pupil, autoregulation in the ipsilateralhemisphere was beginning to fail, and short-lived Cushing responses were making theirpresence evident on the systemic blood pressuretrace. We suggest that the transmission of thepressure wave to the opposite hemisphere there-after resulted in a fall in contralateral hemi-spheric flow, with diencephalic compression andthe fully developed Cushing response. Death frommid-brain compression and medullary ischaemiawould follow shortly thereafter.

Experimental evidence at this time suggeststhat a more gradual rate of expansion of supra-tentorial balloons does not result in the develop-ment of such readily demonstrable differentialpressures. The explanation probably lies in theastonishing viscoelastic properties of the brain,which enable it to change shape while pre-serving function within limits; it seems likelythat this represents a further adaptive factor inthe pressure flow relationship. One would sup-pose therefore that, with a slower rate ofexpansion, the failure of autoregulation andcritical displacements in the brain-stem would bedelayed until a much larger lesion had beenaccommodated. At a critical stage, however, aslight further incremental expansion might resultin the rapid development of differential pres-sures, excessive shifts, and sudden clinicaldeterioration. Nevertheless, it has been shown byDorsch and Symon (1972) and by O'Brien andWaltz (1972) that differential pressures may berecorded even with the relatively slow expansionproduced by infarction of one hemisphere. Thedegree to which the plastic properties of thebrain are capable of evening out intra-tissuepressure gradients is a subject which deservesstudy and which will clearly stimulate methodo-logical advance in the near future.

625



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



REFERENCES

Aukland, K., Bower, B. F., and Berliner, R. W. (1964).Measurement of local blood flow with hydrogen gas.

Circulation Research, 14, 164-187.Aukland, K. (1965). Hydrogen polarography in measure-

ment of local blood flow; theoretical and empirical basis.In Regional Cerebral Blood Flow, pp. 42-45. Acta Neuro-logica Scandinavica, S. 14, pp. 42-45.

Come, S. J., Stephens, R. J., and Symon, L. (1968). Theeffect of drugs on the intracranial pressure of baboons.British Journal of Pharmacology, Proceedihgs of thePharmacological Society, 34, 212P.

Dorsch, N. W. C., Stephens, R. J., and Symon, L. (1971). Anintracranial pressure transducer. Biomedical Engineering,6, 452-457.

Dorsch, N. W. C., and Symon, L. (1972). Intracranialpressure changes in acute ischaemic regions of the primatehemisphere. In Intracranial Pressure, pp. 109-114. Editedby M. Brock and H. Dietz. Springer: Berlin.

H0edt-Rasmussen, K., Sveinsdottir, E., and Lassen, N. A.(1966). Regional cerebral blood flow in man determined byifttra-arterial injection of radioactive inert gas. CirculationResearch, 18, 237-247.

H0edt-Rasmussen, K. (1967). Regional cerebral blood flow.The intra-arterial injection method. Acta NeurologicaScandinavica, 43, Supplement 27.

Hulme, A., and Cooper, R. (1966). A technique for theinvestigation of intracranial pressure in man. Journal ofNeurology, Neurosurgery, and Psychiatry, 29, 154-156.

Jacobson, S. A., and Rothballer, A. B. (1967). Prolongedmeasurement of experimental intracranial pressure using a

subminiature absolute pressure transducer. Journal ofNeurosurgery, 26, 603-608.

Johnston, I. H., Rowan, J. O., Harper, A. M., and Jennett,W. B. (1972). Raised intracranial pressure and cerebralblood flow. I. Cisterna magna infusion in primates. Journalof Neurology, Neurosurgery, and Psychiatry, 35, 285-296.

Johnston, I. H., Rowan, J. O., Harper, A. M., and Jennett,W. B. (1973). Raised intracranial pressure and cerebralblood flow. 2. Supratentorial and infratentorial masslesions in primates. Journal of Neurology, Neurosurgery,and Psychiatry, 36, 161-170.

Johnston, I. H., and Rowan, J. 0. (1974). Intracranial pres-sure gradients in cerebral hemisphere blood flow differencesduring expansion of unilateral supratentorial mass lesionsin primates. In Cerebral Circulation and Metabolism,

Proceedings of the 6th International Cerebral Blood FlowSymposium, Philadelphia, 1973. (In press.)

Langfitt, T. W. (1969). Increased intracranial pressure.Clinical Neurosurgery, 16, 436-471.

Lassen, N. A., H0edt-Rasmussen, K., S0rensen, S. C.,Skinh0j, E., Cronquist, S., Bodforss, B., and Ingvar,D. H. (1963). Regional cerebral blood flow in man de-termined by krypton 85. Neurology (Minneap.), 13, 719-727.

Lundberg, N. (1960). Continuous recording and control ofventricular fluid pressure in neurosurgical practice. ActaPsychiatrica Scandinavica, 36, Supplement 149.

Majors, R., Schettini, A., Mahig, J., and Nevis, A. H. (1972).Intracranial pressures measured with the coplanar pressuretransducer. Medical and Biological Engineering, 10, 724-733.

Nakatani, S., and Ommaya, A. K. (1972). A critical rate ofcerebral compression. In Intracranial Pressure, pp. 144-148. Edited by M. Brock and H. Dietz. Springer: Berlin.

O'Brien, M. D., and Waltz, A. G. (1972). Intracranialpressure changes during experimental cerebral infarction.In Intracranial Pressure, pp. 105-108. Edited by M. Brockand H. Dietz. Springer: Berlin.

Pasztor, E., Symon, L., Dorsch, N. W. C., and Branston,N. M. (1973). The hydrogen clearance method in assess-ment of blood flow in cortex, white matter and deep nucleiof baboons. Stroke, 4, 556-567.

Schettini, A., McKay, L., Majors, R., Mahig, J., and Nevis,A. H. (1971). Experimental approach for monitoring sur-face brain pressure. Journal of Neurosurgery, 34, 38-47.

Sundbarg, G., and Nornes, H. (1972). Simultaneous record-ing of the epidural and ventricular fluid pressure. In Intra-cranial Pressure, pp. 46-50. Edited by M. Brock and H.Dietz. Springer: Berlin.

Symon, L., Dorsch, N. W. C., and Stephens, R. J. (1972).Pressure waves in so-called low-pressure hydrocephalus.Lancet, 2, 1291-1292.

Symon, L., Pasztor, E., Dorsch, N. W. C., and Branston,N. M. (1973). Physiological responses of local areas ofthe cerebral circulation in experimental primates deter-mined by the method of hydrogen clearance. Stroke, 4,632-642.

Tindall, G. T., McGraw, C. P., and Iwata, K. (1972). Sub-dural pressure monitoring in head-injured patients. InIntracranial Pressure, pp. 9-13. Edited by M. Brock andH. Dietz. Springer: Berlin.

626



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Effect ofsupratentorial space-occupying lesions regional ...Effect ofsupratentorial space-occupying lesions on regional intracranial pressure andlocal cerebral blood flow: anexperimental

Documents