sign in sign up
<<
Home , Brain biopsy, Brain herniation

J7ournal ofNeurology, , and Psychiatry 1993;56:845-858 845

NEUROLOGICAL EMERGENCY J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from

Management of raised

J D Pickard, M Czosnyka

Epidemiology The 'four-lump' concept describes most Raised intracranial pressure (ICP) is the final simply the causes of raised ICP: the mass, common pathway for many intracranial prob- CSF accumulation, vascular congestion and lems (table 1) and has a profound influence cerebral oedema (table 2).>7 The description on outcome. For example, of the 3-500 000 of a patient with raised ICP as having cere- patients with head injury seen in Accident bral congestion, vasogenic oedema etc, can and Emergency Departments in the United only be a working approximation, albeit use- Kingdom per annum, 20% are admitted of ful, until our rather crude methods of assess- whom 10% are in (2% of all attenders). ment are refined. In adults, the normal ICP Over 50% of these have an intracranial pres- under resting conditions is between 0-10 sure greater than 20 mm Hg.'2 A total of mmHg with 15 mmHg being the upper limit 80% of patients with fatal head injuries (4% of normal. Active treatment is normally insti- of all patients with head injuries admitted) tuted if ICP exceeds 25 mmHg for more than show evidence of a significant increase in 5 minutes although a treatment threshold of intracranial pressure at necropsy. Some 35% 15-20 mmHg has been suggested to improve of severe head injuries die and 18% are left outcome.8 In the very young, the upper limit severely disabled at enormous financial and of normal ICP is up to 5 mmHg.9 Small emotional cost to the family and community. increases in mass may be compensated for by Similarly, 20 per 100 000 per year are admit- reduction in CSF volume and cerebral blood ted with intracerebral haematoma and 10-12 volume but, once such mechanisms are per 100 000 per annum with subarachnoid exhausted, ICP rises with increasing pulse haemorrhage. The average regional neurosur- pressure and the appearance of spontaneous gical unit serving a population of two million waves (plateau and B waves).10 There is an will manage 200 patients per annum with exponential relationship between increase in tumours, some 15 patients with chronic volume of an intracranial mass and the subdural haematoma, and a similar number increase in intracranial pressure at least of patients with a cerebral abscess and 50 within the clinically significant range.

patients with .3 In comatose Cerebral perfusion pressure is commonly http://jnnp.bmj.com/ children the incidence of raised ICP was 53% defined as: CPP = mean arterial blood pres- of those with head injuries, 23% with anoxic- sure-mean ICP; mean ICP closely approxi- ischaemia damage, 66% with , mates to mean cerebral venous pressure. As 57% with , 100% with mass cerebral perfusion pressure (CPP) falls with lesions and 80% with hydrocephalus.4 There increasing ICP, ICP pulse pressure increases is a considerable risk in all such patients of (fig 1).1 12 Firstly, the brain is less compliant secondary with long term or stiffer and a given pulsatile cerebral blood severe disability if raised intracranial pressure volume load provokes a bigger pressure on October 1, 2021 by guest. Protected copyright. is not recognised and managed appropriately. response. Secondly, the pulsatile component of cerebral blood flow increases with decreas- ing CPP. Cerebral arteriolar dilatation to Pathophysiology maintain cerebral perfusion ('autoregulation') Resting ICP represents that equilibrium pres- may be involved. The lower limit of CPP sure at which CSF production and absorption which will permit autoregulation, when ICP are in balance and is associated with an is raised, is about 40 mmHg. However, there equivalent equilibrium volume of CSF. CSF is a paradox: the level of CPP below which is actively secreted by the choroid plexus at outcome after severe head injury and associ- about 0-35 ml/minute and production ated parameters deteriorate is of the order of remains constant provided cerebral perfusion 60-65 mmHg (MAP < 80 mmHg; ICP > 20 pressure is adequate. CSF absorption is a mmHg). Conventionally any elevation of ICP passive process through the arachnoid requires treatment if CPP is below 60 mmHg Academic and increases with CSF in adults for over 5 minutes. This paradox Neurosurgical Unit granulations rising (Box 167), Level 4, A pressure: may partly reflect the 'split brain' problem: Block, Addenbrooke's autoregulation of CBF to changes in CPP Hospital, Cambridge, CSF drainage = CSF pressure-sagittal sinus pressure and response to changes in arterial carbon CB2 2QQ, UK outflow resistance dioxide tension be J D Pickard (PaCO2) may impaired M Czosnyka (Reference 22). focally, leaving intact reactivity in other areas 846 Pickard, Czosnyka

Table 1 Some common causes ofraised ICP 15 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from Head Injury Intracranial haematoma (extradural, subdural and intracerebral) Diffuse brain swelling

Contusion .: I Cerebrovascular I . Subarachnoid haemorrhage I.' on Intracerebral haematoma E Hydrocephalus E Cerebral venous thrombosis 0L *t~~~I Major cerebral infarct Hypertensive (malignant , eclampsia) Hydrocephalus Congenital or acquired Obstructive or communicating II Craniocerebral disproportion Brain "Tumour" (cysts; benign or malignant tumours) 20 Hydrocephalus 40 Mass effect Oedema 'Benign' Intracranial Hypertension CNS Meningitis Encephalitis 0) Abscess I Cerebral malaria E 2' Hydrocephalus E Metabolic Encephalopathy Hypoxic-ischaemic a- Reyes syndrome etc. Lead encephalopathy Hepatic coma Renal failure Diabetic ketoacidosis Burns Near drowning Hyponatraemia 0 80 ICP (mm Hg) (Adapted from references 4, 5, 6, 7) Figure 1 Relationship between mean ICP and amplitude ofthe ICP waveform in two patients. In the lower trace, there is an upper breakpoint in this relationship when CPP (M;AP-ICP) is less than 30 mmHg. of the brain. If vasospasm is present, an even higher perfusion pressure may be required to provide adequate levels of cerebral blood reason for the treatment of raised ICP by flow. One interesting phenomenon revealed hyperventilation: an acute reduction of by (which reflects flow PaCO, vasoconstricts normal cerebral arteri- in large vessels) and laser Doppler (which oles and thereby directing blood to focally reflects tissue perfusion) is the change in flow abnormal areas. and perfusion during the cardiac cycle: dias- Normally, cerebral blood flow is coupled to tolic perfusion pressure may be below the cerebral oxidative metabolism via multiple normal limit of autoregulation whilst systolic mechanisms involving local concentrations of

is above (vide infra). hydrogen ions, potassium and adenosine for http://jnnp.bmj.com/ Total cerebral blood flow may be increased example. Status epilepticus leads to gross or decreased in areas with absent reactivity. cerebral vasodilatation and intracranial hyper- Hyperaemia is non-nutritional 'luxury perfu- tension as a result of greatly increased cere- sion' where CBF is in excess of the brain's bral metabolism and local release of metabolic requirements13 and accompanied endogenous vasodilator agents. Depression of by early filling veins on angiography and 'red cerebral energy metabolism by anaesthesia veins' at operation. Cerebral vasodilators and hypothermia may reduce cerebral blood such as carbon dioxide will dilate 'normal' flow and ICP where there is a large area of on October 1, 2021 by guest. Protected copyright. arterioles, increase ICP and may run the risk the brain with reasonable electrical activity14 of reducing flow to damaged areas of brain and where normal flow-metabolism coupling (intracerebral 'steal'). Inverse 'steal' is one mechanisms are intact as indicated by a rea- sonable CBF CO2 reactivity.'5 Spontaneous waves of ICP are associated with cerebrovascular dilatation. Cerebral Table 2 Mechanisms ofraised ICP blood volume increases during plateau waves A Mass Lesions Haematoma, abscess, tumour. (ICP > 50 mmHg for more than 5 minutes) B CSFAccumulation Hydrocephalus (obstructive and communicating) and including contralateral ventricular dilatation from and may be the result in some cases of inap- supratentorial brain shift. propriate autoregulatory vasodilatation in C Cerebral Oedema Increase in brain volume as a result of increased water content. response to a critical fall in CPP but certainly 1 Vasogenic-vessel damage (tumour, abscess, not in all cases (fig 2).16 TCD has revealed contusion) 2 Cytotoxic-cell membrane pump failure that middle cerebral artery (MCA) flow (hypoxaemia, ischaemia, toxins; myelinoclastic). velocity increases pari-passu with B waves 3 Hydrostatic-high vascular transmural pressure (loss of autoregulation; post intracranial decompression). (0-5-2/min) of ICP (fig 3).17 4 Hypo-osmolar-hyponatraemia Finally, gradients of ICP may develop 5 Interstitial-high CSF pressure (hydrocephalus) D Vascular (congestive) brain Increased cerebral blood volume when herniation occurs-transtentorial, sub- swelling -arterial vasodilatation (active, passive) falcine and . Blockage to -venous congestion/obstruction. the free flow of CSF between intracranial Management ofraised intracranialpressure 847

Figure 2 Relationships Brain death J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from between arteral pressure (ABP), CPP ICP amplitude (MP) during AMP116 a plateau wave. (mm Hg) Km- 01:40 40 ^ | E

ICmPm g 34-85 U-

120 ABP 86-23 (mm Hg) Figure 4 Reversal of middle cerebral arteryflow velocity (FV) in a patient whofudfils the criteria for T| ~~~~~~~~01:40 death.

CPP513 (mm Hg) _ 13 30 Monitoring techniques 1 hour A) Clinicalfeatures In the non-trauma patient, there may or may compartments leads to a much greater and not be a clear history of , vomiting more rapid rise in ICP in the compartment and visual disturbance suggestive of papil- harbouring the primary and hence loedema or a VIth nerve palsy. The absence to the final common sequence of transtentor- of papilloedema does not exclude raised ICP ial and foramen magnum coning. When ICP in patients with acute or chronic problems: equals arterial blood pressure, angiographic disc swelling was found in only 4% of head pseudo-occlusion occurs, reverberation, sys- injury patients, 50% of whom had raised ICP tolic spikes or no flow may be seen on TCD on monitoring.6 Even in the 1 990s, it is (fig 4). Patients will often satisfy the formal regrettable that a clear history of raised clinical criteria for brainstem death, for which intracranial pressure may be misinterpreted TCD is not a substitute.1819 until the final denouement of disturbance of There is a complex interaction between the and pupillary abnormality or properties of the CSF and the cerebral circu- apnoea presents. Only slowly has the danger lations that may be modelled (fig 5).20 21 The of in the differential diagno- relative contributions of abnormalities of CSF sis of neurological patients been appreciated absorption and cerebral blood volume may be by the non-expert. Many of the later signs of approximated by calculating the proportion raised ICP are the result of herniation: moni- of CSF pressure attributable to CSF outflow toring should detect raised ICP at an earlier resistance and venous pressure from Davson's stage and hence treat before irreversible equation (ICP = CSF formation rate x out- damage occurs. flow resistance + sagittal sinus pressure).22

Phenomena such as the interaction of B) CT scanning http://jnnp.bmj.com/ autoregulation to changing CPP with PaCO2 CT scanning may reveal not only a mass, may be quantified.2' hydrocephalus or cerebral oedema but also evidence of diffuse brain swelling such as absent perimesencephalic cisterns, com- pressed 3rd ventricle and midline shift. 30 ICP 22-4 C) Invasive methods ofICP monitoring (mm Hg) - on October 1, 2021 by guest. Protected copyright. 15_ including CSF infusion tests The gold standard of ICP monitoring, that 4 was first introduced between 1951,2410 still 95 86.8 remains the measurement of intraventricular ABP or a CSF (mm Hg) fluid pressure either directly via 85 reservoir, with the opportunity to exclude zero drift. Subdural fluid filled catheters are 45 reasonably accurate below 30 mm Hg. A total FVx 35-6 of 25 mmHg for more than 5 minutes is the (cm/s) treatment 30 usual threshold level at which should be instigated. Risk of infection, and haemorrhage is less with sub- 115 dural than intraventricular catheters but even HR 111-8 the latter should be less than 5% overall. (b/m) transducers are very useful par- 105 Catheter tip 113:56 ticularly for waveform analysis, whether placed intraventricularly, subdurally or 1 minute Ventricular catheters similr.intracerebrally. permit Figure 3 B waves ofICP in a head-injured patient and their relationship to the of CSF in cases of variations in middle cerebral arteryflow velocity compared to fluctuations in arterial therapeutic drainage pressure (ABP) and heart-rate (HR). ventricular dilatation. In more chronic condi- 848 Pickard, Czosnyka

Extracranial tions of ventricular dilatation, where ICP is J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from cerebral r- not greatly raised, obstruction to CSF Saggital absorption may be confirmed by CSF infu- slnus sion tests (ventricular or lumbar) taking care ("P"') to adapt the technique to the site of any obstruction.2527 Twenty four hour intracra- nial pressure monitoring in patients with so- called normal pressure hydrocephalus may reveal a high incidence of B waves during sleep which is a very helpful prognostic sign for the outcome following shunting.2829 Benign intracranial hypertension seldom requires more than CSF pressure monitoring through a lumbar catheter or needle for an hour. Considerable effort continues into the close analysis of the ICP trace to determine whether it is possible to reveal the mechanism of raised ICP and whether autoregulatory reserve remains intact.30 It has been proposed7 that congestion or vascular brain PSS swelling may be present when the ratio of the amplitudes of the pulse and respiratory com- ponents of the ICP trace exceeds 2, when there is an increase in the high frequency cen- troid or when there is a high amplitude trans- fer function for the fundamental harmonic from ABP to ICP. Such a transfer function is calculated from the first Fourier transform of Figure 5 Hydrodynamic model ofcerebral bloodflow and CSF circulation w;,ith the the digitised signal. However, continuous electrically equivalent circuit (for details, see Czosnyka, et al, reference 21). multimodality monitoring is required to draw any safe conclusions and should include some measure of CBF (for example, TCD) and cerebral metabolism (for example, EEG, jugular venous oxygen). Indices of imminent decompensation would be very helpful but volume-pressure responses,31 pressure-volume indices or definition of the contribution of CSF outflow resistance to ICP,22 are not suit- able for routine clinical use. Figure 6 Relationship 70 between ICP, CPP, ICP ICP D) Non-invasive ICP monitoring amplitude (AMP) and (mm Hg) flow velocity (systolic FV, 20 It would be very helpful to monitor ICP or http://jnnp.bmj.com/ diastolic FVd and hi. CPP without invasive catheters. Transcranial amplitude FVJ during a 70 i plateau wave. The CPP Doppler, tympanic membrane displacement amplitude ofICP increases (mm Hg) and even compliance studies have been with ICP and decrease in 20 advocated. It would be very helpful to have CPP: flow velocity amplitude increases as a 30 answers to the following questions: What is result of the fall in diastolic AMP the cerebral perfusion pressure at any given flow velocity. (mm Hg) time? What is the relative contribution of

0 on October 1, 2021 by guest. Protected copyright. each possible mechanism to raised ICP? 80 i FVa What features may predict decompensation? (cm/s) Is it possible to have an on-line assessment of 50 I cerebrovascular reactivity either to changes in 120 CPP (autoregulation) or CO2? Which therapy CPPs * or cocktail is best suited to the sum of that (mm Hg) * individual's 'split-brain' problems? 0 fl A noninvasive method that monitored con- 60 tinuously both CPP and CBF autoregulatory CPPd reserve would be very helpful in refining man- (mm Hg) 0 _ agement of the swollen brain. 140 FVs TRANSCRANIAL DOPPLER (cm/s) Aaslid's description of transcranial Doppler 0 in 1982 permitted bedside monitoring of one 50 index of CBF, non-invasively, repeatedly, and FVd even continuously.32 33 The problem has been (cm/s) 0 that it is a big tube technique that measures flow velocity in branches of the Circle of 1 minute Willis, most commonly the middle cerebral Management ofraised intracranial pressure 849

artery. Changes in velocity may reflect either relate with very severe brain injury or with J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from changes in blood flow or in diameter of the extensive focal lesions in MCA distribution. insonated artery. Unfortunately, diameter Knowledge of the integrity of the CBF CO2 and flow may not change in complementary response was helpful in determining the directions and great care must be taken with potential effectiveness of hyperventilation or the interpretation.34 Low velocity may indi- barbiturates for ICP control.40 cate low flow or arterial dilatation at constant flow. High velocity may indicate high flow or TYMPANIC MEMBRANE DISPLACEMENT. arterial constriction/vasospasm at constant ICP is transmitted via the cochlear aqueduct flow. Considerable ingenuity has been to the perilymph of the cochlea providing that expended in analysis of a TCD wave form. the aqueduct is patent. Perilymphatic pres- The amplitude of the flow velocity pulse wave sure may be assessed indirectly by recording (FVa) reflects pulsatile changes in regional displacement of the tympanic membrane dur- CBF and is dependent on the amplitude of ing stapedial reflex contractions elicited by a the arterial pressure wave, regional cere- loud sound.4' High perilymphatic pressure brovascular resistance, the elastance of the displaces the resting position of the stapes capillary bed and the basal cerebral arteries. footplate laterally, thereby allowing a higher Aaslid suggested that an index of CPP could degree of motion in a medial direction, and be derived from the ratio of the amplitudes of results in a more inward going tympanic the first harmonics of the arterial blood pres- membrane displacement on stapedial con- sure and of the middle cerebral artery velocity traction. Low perilymphatic pressure will (detected by TCD) multiplied by mean flow have an opposite effect. A transducer probe velocity. There is a reasonable correlation attached to a head set is placed in the between the pulsatility index (peak systolic- patient's external auditory meatus and com- end diastolic FV's/time averaged FVm) of puter based instrumentation allows small MCA velocity and CPP after head injury but movements of the tympanic membrane to be absolute measurements of CPP cannot be measured when 100OHz of increasing sound extrapolated.35 Nelson et aP6 have provided pressure level induces stapedial contraction. both experimental and theoretical modelling This very ingenious technique is useful in evidence for three haemodynamic phases as younger patients with hydrocephalus or CPP falls. Above the lower limit of autoregu- benign intracranial hypertension on a sequen- lation, falls in CPP are masked by arteriolar tial basis, provided that a skilled audiologist is dilatation, a fall in CVR and gradual increase available. It does not provide an absolute in FVa so that CBF and FVm remain stable. measure of ICP. It is of no value in patients During the transitional phase, CBF and FVm on ventilators receiving muscle relaxants. The start to fall gradually: CPP in diastole is close patency of the cochlear aqueduct decreases to or below the critical closing pressure of the with age and should be checked with a pos- capillaries so that the fall in FV in diastole is tural test. greater and FVa increases further. Finally autoregulation becomes exhausted with a E) Cerebral venous oxygen42 43 rapid fall in CBF and FVm a sharp decrease in Cerebral arterio-venous oxygen content dif- FVa and an increase in CVR. Where autoreg- ference should normally be 5-7 ml/dl. Values

ulation is impaired throughout, CBF and below 4ml/dl indicate cerebral hyperaemia http://jnnp.bmj.com/ FVm fall pressure-passively as CPP is reduced while values above 9ml/dl indicate global and there is no increase in FVa. The correla- cerebral ischaemia. Jugular bulb oxygen satu- tion coefficient between FV. and ABP may ration may be monitored, preferably continu- provide a continuous on-line index of ously, with an indwelling catheter. Single autoregulatory reserve: as autoregulation measurements of jugular venous oxygen are becomes exhausted, correlation rapidly moves of little value given the many fluctuations to The response of the during the day. Overenthusiastic treatment,

from negative positive. on October 1, 2021 by guest. Protected copyright. cerebral circulation to stress, such as a period that on occasion may induce cerebral of hypotension, hypercapnia or transient ischaemia, may be monitored with this tech- carotid compression may also be assessed."21 nique. Hyperventilation and barbiturate- Hence, autoregulation and cerebrovascular induced falls in CPP have been shown in reactivity may be assessed within the vascular individual patients to be counter-productive. territory supplied by the insonated artery. An index of regional oxygen metabolism is Comparison of the changes in blood flow required. Transcutaneous, transcranial near velocity in a number of branches of the Circle infra-red spectroscopy is completely non- of Willis with that in the cervical internal invasive.4 Unfortunately, it is well proven carotid artery and with cerebral arterio- only in neonates and younger children and venous oxygen difference may help to distin- not yet in older age groups. guish vasospasm from hyperaemia at least on a global basis. Vasospasm defined by TCD F) Cerebral electical activity has been associated with delayed cerebral The compressed EEG (cerebral function ischaemia after trauma.'7-'9 Such techniques monitoring) is helpful in deciding whether cannot yet define the proportional contribu- cerebral metabolic depressants may be indi- tions of both vasospasm and hyperaemia in cated in the treatment of intracranial hyper- the same or different parts of the brain in tension.'4 Such drugs will obviously not be patients who may have both. Depressed CBF helpful if the EEG is flat or greatly reduced in CO, reactivity was found in one study to cor- amplitude. 850 Pickard, Czosnyka

Table 3 Indications for ICP monitoring between ICU and CT scanning too often J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from Head injury involves well documented risks of hypoxia (a) being artificially ventilated: and hypotension; ICP should not be -coma with compression of 3rd ventricle and/or therapy reduction in perimesencephalic cisterns on CT, used as a blunderbuss but needs to be selec- -coma following removal of intracranial haematoma, tively targetted if it is not to be counter-pro- -coma with abnormal motor response as the best reaction, ductive, and new treatments are emerging -coma with midline shift/unilateral ventricular Clinical trials based on outcome dilatation, very rapidly. -multiple injury including severe chest wall injuries. studies at six months in such heterogeneous -early not easily controlled. groups of patients may easily miss useful -refractory hyperpyrexia. benefits.45 To treat raised it must first be (b) uncertainty over surgery for small haematoma/multiple ICP, lesions. identified, avoidable factors prevented or Intracerebral and subarachnoid haemorrhage treated and finally active treatment instigated, -coma based on our -postoperatively following intraoperative complications, hopefully understanding of the -hydrocephalus. individual mechanism involved. ICP should Coma with brain swelling be treated before herniation occurs-clinical -metabolic, -hypoxia/ischaemia, signs particularly in patients on a ventilator -infective (see Table 1). are just too crude. Knowledge of ICP may Hydrocephalus and benign intracranial hypertension help with prognosis and counselling of rela- (Adapted from references 4, 7, 42) tives: in one series of diffuse head injuries, where ICP persistently exceeded 20 mmHg, almost all the patients died compared with a Management strategies mortality rate of 20% in those where ICP 1) Emergency resuscitation and diagnosis could be kept below 20mmHg with treat- Patients who are rapidly deteriorating or ment.46 already in coma require immediate resuscita- tion if necessary with intubation and ventila- 3) Prevention ofintracranial hypertension: tion followed by a diagnostic CT scan. An general medical and nursing care-avoidable intravenous bolus of mannitol (0-5 gms/kg factors. over 15 minutes may be required if there is Simple measures need to be checked (table evidence of coning such as pupillary dilata- 4).7 42 Ideally the position of the patient tion.) Acute ventricular dilatation demands should minimise any obstruction to cerebral immediate ventricular drainage-bilateral if venous drainage by head-up tilt whilst avoid- the lesion is midline. Hyperacute ventricular ing any fall in cardiac output or carotid arter- dilatation following subarachnoid haemor- ial blood pressure. Direct measurement of rhage or in association with a third ventricular global CBF and CPP suggests that head-up lesion need not be gross to cause death. tilt of up to 30° is safe but careful scrutiny Surgical clots require removal and abscesses should be kept of CPP in individual require tapping. patients.47-50 Hypovolaemia should be avoided, contrary 2) Post-acute management to some historical teaching. The evidence is Many neurosurgical units worldwide still most clear cut in patients after subarachnoid manage patients without the help or hin- haemorrhage: dehydration particularly when

drance of ICP monitoring. The nihilistic coupled with hyponatraemia increases the http://jnnp.bmj.com/ argument is that ICP monitoring has not risk of cerebral infarction.51 Patients with CT been clearly shown to improve outcome and evidence of raised ICP are already at greater sequential CT scanning provides sufficient risk of hypovolaemia after SAH.52 information. Treatment for raised ICP has Dehydration increases the risk of hypo- not greatly advanced over the past two volaemia which may be revealed only when decades and can be applied pragmatically. the patient is given an anaesthetic agent, for The alternative school of thought argues that example, for an orthopaedic procedure or as ICP monitoring should be selective (table 3) part of a regime to control raised ICP. A on October 1, 2021 by guest. Protected copyright. based in part on the initial CT scan. stable circulation must be maintained,53 if However, in addition such monitoring is very necessary with colloid and inotropes (dobuta- educational and greatly assists general nurs- mine or dopamine for its renal sparing ing and medical care; transport of a patient action). However, overenthusiastic hyperten- sive-hypervolaemic therapy remains very con- troversial in the context of head injury with its Table 4 Potential problems exacerbating raised intracranial pressure multiple pathology and uncertainty over the integrity of the blood-brain barrier.5456 1 Calibration of ICP and arterial blood pressure transducers and monitors particularly to check the zero reference point. Systemic hypertension should not be treated 2 Neck vein obstruction -inappropriate position of head and neck-avoid constricting directly with agents such as sodium nitro- tape around neck. 3 Airway obstruction -inappropriate PEEP, prusside. Sodium nitroprusside impairs -secretions, bronchospasm etc. autoregulation and increases the risk of 4 Inadequate muscle relaxation-breathing against ventilation, -muscle spasms. boundary zone infarction.57 The cause of 5 Hypoxia/hypercapnia hypertension, for example, pain or retention 6 Further mass lesion-rescan 7 Incomplete analgesia, incomplete sedation and anaesthesia. of urine, should be looked for. 8 Seizures The majority of neurosurgical patients with 9 Pyrexia 10 Cerebral vasodilating drugs hyponatraemia do not have inappropriate 11 Hypovolaemia secretion of ADH and it is unwise to use fluid 12 Hyponatraemia (often iatrogenic fluid overload) restriction to treat them even if they do.51 58 Management ofraised intracranial pressure 851

A useful dictum in neurosurgery is that blood molecular weight tracers. The time course is J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from volume comes before plasma sodium levels. slow and does not account for the immediate Moderate hyponatraemia impairs cerebro- effect of mannitol on ICP. In patients with vascular reactivity experimentally to both peritumoural oedema, mannitol causes with- hypercapnia and hypotension but does not drawal of water mainly from brain areas augment the cerebrovascular effects of experi- where the barrier is impaired as judged by Ti mental SAH.59 A spectrum of abnormalities MRI and in vitro measurements of brain give rise to hyponatraemia following SAH, for water content.68 However, mannitol may example: initial natriuresis leading to volume accumulate in oedematous white matter with depletion, ADH secretion stimulated both by repeated doses.69 stress and volume depletion, ADH-induced The more immediate effects of intravenous water retention, steroid- and sympathetic- mannitol include a fall in whole blood viscos- induced effects on the kidney and possible ity with reduced red cell rigidity and corpus- release of atrial natriuretic factor (both car- cular volume, an increase in brain compliance diac and cerebral in origin) and digoxin-like and possibly cerebral vasoconstriction.7>7' substance. Experimental perivascular administration of Seizures may be difficult to recognise when mannitol evokes vasodilatation. The cerebral the patient is paralysed and ventilated. vasoconstriction with intravascular bolus Episodes of pupillary dilatation with increases administration was short-lived-in the cat, in arterial blood pressure and ICP are sugges- both pial arteriolar diameter and ICP tive. returned to normal within 30 minutes and Pyrexia not only increases cerebral metabo- thereafter both increased, pari-passu with lism and hence cerebral vasodilatation but changes in blood viscosity.72 Administration also cerebral oedema. Severe hypothermia over 15 minutes produced no change in pial was used historically to treat raised ICP but it arteriole or venular diameter in another has become clear more recently that mild study.73 Current studies in patients are using hypothermia of a few °C only will reduce transcranial and laser Doppler to re-examine cerebral ischaemia for reasons that are not yet these conflicting reports. Why should a sud- clear."13 Hyperglycaemia should be avoided. den change in blood viscosity evoke acute There is considerable evidence that cerebral transient vasoconstriction? Chronic changes ischaemia and infarction is made worse by in blood viscosity by plasma exchange, with- hyperglycaemia and the use of high glucose out alterations in haemoglobin or arterial oxy- solutions is contraindicated unless there is gen content do not change steady-state CBF significant evidence of benefit in a particular in humans.74 Patients with high plasma vis- metabolic encephalopathy.6061 cosity or with high viscosity due to large numbers of white cells do not have low CBF values. In a series of patients with haemato- Osmotic diuretics logical diseases but no evidence of cere- Intravenous mannitol is invaluable as a first brovascular disease, arterial oxygen content aid measure in a patient with brain herniation was the major determinant of CBF-blood as a result of raised ICP. Its more prolonged viscosity per se had no significant effect on use and mechanisms of action remain con- CBF.75 If a single is considered, tentious issues. Osmotherapy began experi- the apparent viscosity of blood diminishes in http://jnnp.bmj.com/ mentally with hypertonic saline and then proportion to its radius as a result of the mar- urea, entering neurosurgical practice with ginal sheath of low viscosity and axial flow of 30% urea in 1958. A maximum fall in ICP red cells.76 The width of this sheath is greatest occurs within 30 minutes of starting urea, relatively in small vessels. Furthermore, and the effect lasts for up to three hours but apparent viscosity increases with falling veloc- with the possibility of subsequent rebound.62 ity. Hence with pial arterial dilatation, local

Conceptually, the mechanism was thought to blood viscosity will rise both because of the on October 1, 2021 by guest. Protected copyright. be osmotic extraction of water across the increased proportion of red cells and as a intact blood-brain barrier acting as a semiper- result of the reduction in flow velocity if tis- meable membrane. Experimentally, urea sue perfusion flow remains constant. shrank normal rather than oedematous Simplistically, according to Poiseuille, as vis- brain.63 Entry of urea into the oedematous cosity is reduced deliberately, so the pressure brain through a 'defective' barrier would take gradient along the pial arteriole under obser- water in and thereby account for rebound. vation falls. Hence, the distal intravascular Overenthusiastic bolus administration of an pressure increases if the proximal pressure osmotic diuretic may cause abrupt systemic remains constant. The distal end of the arteri- hypertension, an increase in cerebral blood ole therefore constricts if autoregulatory volume if autoregulation is defective or its mechanisms such as the Bayliss effect are upper limit is exceeded and promote - intact. 'Viscosity' autoregulation should tion rather than the reverse.64 depend on pressure autoregulation unless More recent studies indicate that mannitol, there is a separate endothelial mechanism given time, removes water from both normal that is flow- or viscosity-sensitive. and oedematous brain, be it ischaemic or Alternatively, mannitol may transiently interstitial (Marmarou's infusion model).6>7 increase CBF, increase oxygen delivery and The oedematous area around many mass wash out local vasodilators such as lesions may still have an intact blood-brain adenosine.7' Vasoconstriction then follows. barrier at least to the conventional high Extracellular hyperosmolarity is a potent 852 Pickard, Czosnyka

cerebral vasodilator and it is remarkable that hyperventilation may precipitate cerebral the intravenous vasoconstrictor effect of man- ischaemia with EEG slowing, CSF lactic aci- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from nitol so completely dominates the acute cere- dosis and a cerebral arterio-venous oxygen brovascular effect. If the viscosity mechanism content difference > 9 ml/dl. Arterial blood is apposite, it will depend upon the distribu- pressure may be reduced by the combination tion gradient of intravascular pressures along of dehydration and aggressive hyperventila- the cerebrovascular tree which may not be tion. Finally, weaning from the ventilator may easy to predict with different and be more difficult and prone to "rebound" cerebral perfusion pressures. Certainly the intracranial hypertension: the brain's buffer- reported effects of mannitol on CBF are not ing mechanisms have to readjust back again. easy to rationalise.77-80 In severely head Finally, cerebrovascular reactivity to CO2 injured patients in whom autoregulation was may be absent in some patients after head absent, intravenous mannitol caused an injury, but such reactivity is seldom measured increase in cerebral blood flow and no reduc- before hyperventilation is started. TCD is a tion in ICP.8' ICP was reduced in those simple way of assessing CO2 reactivity. In one patients where autoregulation was intact. study, controlled hyperventilation used pro- However, in patients with unruptured phylactically did not improve outcome but aneurysms and in the majority of whom did prolong the recovery phase.88 Hence there autoregulation was intact presumably, manni- is growing awareness that hyperventilation be tol (bolus or infusion) increased CBF for used sparingly, for example, to treat persis- many hours.80 More regional assessments of tent ICP waves. CBF suggest that mannitol may stablise pH CSF lactate accumulation and CSF acido- and CBF in regions of moderate but not sis occur after head injury.89 90 Both severity of severe ischaemia.8' Other suggested mecha- injury and the proportion of patients with a nisms for the effect of mannitol include poor outcome are related to high and increas- movement of water from CSF into capillaries ing CSF lactate levels. Cerebral tissue lactic and scavenging of free radicals.8' Plasma acidosis is related to secondary brain damage hyperosmolality rapidly reverses the intersti- following a primary insult such as cerebral tial fluid pressure/CSF pressure gradient and ischaemia even if moderate acidosis per se has there is a rapid volume shift within 30 min- no persisting effect on normal neurons. utes from CSF into brain tissue."'3 Akiota et a19' found that the intravenous Many attempts have been made to ratio- buffer tris hydroxymethyl aminomethane nalise how much mannitol may be given and (THAM) ameliorated both the CSF acidosis when, for more prolonged effects.8485 In prac- and brain swelling following epidural balloon tice, mannitol tends to be given as an inter- compression of the brain in dogs. In 1970, mittent bolus whenever the individual Gordan and Rossanda92 suggested that hyper- patient's ICP rises significantly above the ventilation might be beneficial as the result of threshold of 25-30 mmHg. The effects of compensation of CSF acidosis but only at mannitol may be potentiated by adding very low arterial pCO2 (20-25 mmHg) that is frusemide.85 It is crucial to avoid dehydration now proposed to produce severe cerebral and latent hypotension with careful attention vasoconstriction and in some individuals to fluid balance. Colloid with an adequate cerebral ischaemia. THAM, after intravenous

plasma half life (albumin, hetastarch for administration, equilibrates with the intracel- http://jnnp.bmj.com/ example) should be combined with careful lular and extracellular spaces in the body as electrolyte replacement. Another dose of well as with CSF. Evidence is accumulating mannitol should not be given if osmolarity both experimentally and in humans that exceeds 330 mmol/l for fear of tubular THAM is at least as effective as mannitol in damage and renal failure. Repeated doses of reducing experimental oedema in the brain mannitol should not be given unless an and lowering ICP after head injury.93 94 ICP monitor is in place. Some authors THAM reduces the demand for mannitol and continue to recommend glycerol for pro- CSF drainage. In the most recently published on October 1, 2021 by guest. Protected copyright. longed osmotherapy.86 randomised prospective clinical trial,94 a total of 149 patients with severe head injury ( < 8) were randomly Hyperventilation, the buffer THAM and assigned to either a control or a THAM indomethacin group. Both groups of patients matched in The cerebral vasoconstrictor effect of terms of clinical parameters including age, hypocapnia, induced by hyperventilation, sex, number of surgical mass lesions, number does not persist much beyond a day, probably in each Glasgow scale stratum and the first in part because the bicarbonate buffering ICP measurement. All patients were treated mechanisms within the brain and cerebrovas- by standard management protocols, intu- cular smooth muscle themselves readjust to bated, mechanically ventilated, and main- return extracellular and intracellular pH tained in the PaCO, range of 32-35 mm Hg nearer to the original values.87 This phenome- for 5 days. THAM was administered as a 0-3 non has now been confirmed in vivo in nor- M solution in an initial loading dose (body mal subjects by magnetic resonance weight x blood acidity deficit, average 4-27 spectroscopy. Neurosurgical patients with cc/kg/hour) given over two hours, followed by healthy lungs and systemic circulation often constant infusion of lml/kg/hour for five days. hyperventilate spontaneously down to a Outcome was measured at three, six and 12 PaCO, of 30 mmHg.8 More enthusiastic months post injury. Although analysis indi- Management ofraised intracranial pressure 853

cated no significant difference between these drained gradually against a positive pressure two groups at three months, six months or of 15-25 cm H,O to avoid unrestrained J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from one year, there was a difference regarding drainage. In the case of a posterior fossa ICP. The time that ICP was above 20 mmHg tumour, upward coning may be precipitated in the first 48 hours after injury was less in if the supratentorial ventricles are drained too patients treated with THAM. Also the num- precipitately. In patients with a hemisphere ber of patients requiring barbiturate coma mass causing midline shift and contralateral was significantly less in the THAM group hydrocephalus, drainage of that ventricle may (5-5% versus 18-4%). The authors concluded make the shift worse. In patients with diffuse that THAM ameliorated the deleterious brain swelling, the ventricles are small and effects of prolonged hyperventilation, was not always easy to cannulate. Stereotactic beneficial in ICP control and further study of techniques are useful but not appropriate in dose and timing of administration was an emergency. Even where ICP is controlled warranted. by drainage against a pressure 15-25 cms The cerebrovascular response to hyper- H2O, such a ventricular catheter readily capnia may be manipulated in other ways. It becomes blocked and is seldom a satisfactory has been known since 1973 that the cere- technique per se. CSF drainage alone is the brovascular CO2 response is blocked by optimal method of controlling intracranial indomethacin in doses that partly inhibit hypertension in patients with subarachnoid brain cyclo-oxygenase activity in vivo.95 haemorrhage where the cause is often distur- Cerebral venous pressure is very significantly bance of the CSF circulation but there is reduced suggesting that ICP is reduced. That probably an increased risk of rebleeding. CSF observation was not used clinically because of drainage is used as a diagnostic technique to fear that indomethacin was inhibiting produc- assess patients in the poorer SAH grades. tion of prostacyclin and that might be coun- When they improve early surgery should be terproductive. Certainly, cerebral oxygen considered. delivery is seriously impaired when Removal of bone flaps or subtemporal indomethacin is given to very young animals decompressions are performed much less fre- or to very pre-term infants undergoing treat- quently nowadays." Patients with large ment for patent ductus arteriosus.96 However, meningiomas may have a smoother post- in five patients with injury with cerebral con- operative course if a flap is removed electively tusion and oedema in whom it was not possi- at the end of the operation rather than as an ble to control ICP by hyperventilation and emergency a few hours later. Benign intracra- barbiturate sedation, indomethacin (bolus nial hypertension can be treated by a combi- injection of 30 mgs followed by 30 mg/hour nation of optic nerve sheath fenestration and for 7 hours) reduced ICP below 20 mmHg lumbar peritoneal or cisterno-peritoneal for several hours.97 CBF was reduced at two shunting: subtemporal decompressions are hours without any changes in cerebral arteri- rarely indicated. Babies with complex forms ovenous oxygen or lactate differences. Rectal of craniosynostosis may require craniofacial temperature also fell from 38-6 to 37T3°C. surgery to expand the volume of the skull. Slit Hence a more substantial trial of ventricle syndrome for shunt-induced CSF indomethacin appears warranted, perhaps overdrainage may be managed by use of

avoiding young children until further experi- siphon control devices or programmable http://jnnp.bmj.com/ ence is accumulated. Inhibition of nitric oxide valves-subtemporal decompression is sel- synthesis also blocks the cerebrovascular CO, dom required. This procedure was sometimes response but may also increase focal cerebral followed by epilepsy. infarction in the rat middle cerebral occlusion There is a very restricted place for decom- model. pressive following head injury and there is the potential to do considerable Continuous CSF drainage and surgical harmn.9'00 With a very tight brain, opening decompression the dura induces herniation through the on October 1, 2021 by guest. Protected copyright. External ventricular drainage via a catheter or defect. Cerebral venous drainage from the reservoir is a rapid procedure in an emergen- herniated brain obstructs and further brain cly in a patient with hydrocephalus. swelling ensues with infarction. Experi- Biventricular drainage is required for 3rd ven- mentally, craniectomy facilitates formation of tricular lesions that occlude both Foramena hydrostatic brain oedema as might be of Munro. Colloid cysts are best dealt with as expected from consideration of Starling's the primary procedure unless the patient is in equation.'01 Craniectomy may be considered extremis. Patients with communicating in young patients without evidence of diffuse hydrocephalus or benign intracranial hyper- axonal injury (high Glasgow Coma Score on tension may be temporarily controlled by admission) and evidence of diffuse swelling.'02 lumbar drainage through an indwelling For paediatric , Kirkham catheter. It is unkind, unnecessary and less makes a case for performing decompression effective to use repeated lumbar punctures. It earlier rather than later and certainly before is becoming recognised that permanent CSF the EEG disappears.'03 drainage via lumbar peritoneal shunts may be complicated by secondary descent of the Steroids, free radical scavengers and the cerebellar tonsils in patients with no previous lazaroids evidence of a .98 In all The mechanism of the remarkable effect of cases of external drainage, CSF should be glucocorticoids such as dexamethasone on 854 Pickard, Czosnyka

focal, relatively chronic cerebral lesions ethylene glycol conjugated superoxide dismu- remains incompletely understood. Patients tase-have recently been reported but a much J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from deteriorating with a cerebral tumour or an larger trial is required to establish efficacy."12 abscess rapidly improve within 24 hours. It is not yet proven whether steroids help trau- matic cerebral contusions. ICP waves and Cerebral metabolic depressants: compliance improve together although mean excitotoxic amino acid antagonists ICP and water content take days longer to The cerebral metabolic depressant effect of subside. Brain for tumour is much deep hypothermia is now seldom used except safer after at least three days of dexametha- during cardio-pulmonary bypass and total cir- sone (10-20 mgs loading dose; 4 mgs four culatory arrest. Such a technique may be times a day thereafter) particularly when used for complex basilar aneurysms where combined with stereotactic biopsy tech- interventional radiological techniques are niques. Care should be taken to counsel inappropriate. Brain energy metabolism is patients about the side effects of steroids even depressed more conveniently by hypnotic with short courses. agents including barbiturates, etomidate, Much controversy has surrounded the use propofol, althesin and gamma hydroxybu- of very high dose steroids in head injury but tyrate. Unfortunately all such agents have careful controlled trials have shown no bene- side effects, the most relevant of which is sys- fit and in one study the outcome of the treat- temic hypotension, often compounded by ment group was worse.'04-108 However, even dehydration or hypovolaemia. It is essential higher doses started within a few hours of to maintain a normal arterial blood pressure injury are currently under scrutiny. This and not allow CPP to fall. Central venous rationale is based on trials in pressure monitoring is required. One factor injury of methylprednisolone (30 mgs/kg/day) maintaining CPP in patients with raised ICP which showed a modest benefit in the group may be the Cushing mechanism-lower the where it was started within eight hours of ICP and CPP falls. injury. Hypnotic agents depress cerebral oxidative One purported mechanism of action for metabolism and hence lower CBF, CBV and steroids involves lipid peroxidation and free ICP. However, cerebral electrical activity and radicals.'09110 Oxygen is needed for aerobic normal coupling mechanisms between metab- life but it has toxic properties. All organisms olism and flow must be present if barbiturates are subject to oxidative stress as up to 2% of are to lower ICP.14'542 Normal flow metabo- oxygen consumed by the brain, for example, lism coupling mechanisms may be assessed is used to form semi-reduced oxygen interme- by the cerebrovascular response to CO2. diates: superoxide, hydrogen peroxide and Short-term protection during aneurysmal hydroxyl free radicals. These may be used as surgery with barbiturate or propofol is widely part of normal biochemistry or if the safety used. Synergy with even moderate hypother- mechanisms fail-superoxide dismutase, mia may be helpful provided MAP is main- catalase, glutathione peroxidase, glutathione, tained.42 After initial reports of the vitamin E and ascorbate-then such reactive effectiveness of short acting barbiturates in oxygen species may attack nucleic acids, pro- lowering ICP after head injury, three con- teins, carbohydrates and particularly lipids in trolled trials have failed to show any overall the brain. Ferrous iron from blood clots is significant improvement of outcome or http://jnnp.bmj.com/ also active along with such reactive oxygen reduction in number of patients dying with species. Cerebrovascular effects of acute intracranial hypertension."11l16 Such trials hypertension and subarachnoid haemorrhage involved heterogeneous groups of patients, may involve free radical mechanisms damag- however, and a treatment benefit in a sub ing the endothelium. Non-glucocorticoid group may have been missed. steroid analogues of methylprednisolone as In the UK, althesin has been withdrawn

well as methylprednisolone itself weakly even though its idiosyncratic allergic prob- on October 1, 2021 by guest. Protected copyright. inhibit lipid peroxidation. The 21-amino- lems would not have been a contra-indication steroids (the antioxidant family known as the in the intensive care environment. Etomidate lazaroids) are potent inhibitors of lipid perox- blocks steroidogenesis but it is apparently still idation and have a vitamin E sparing effect."' used in the USA as an intraoperative protec- Various experimental models of head and tion agent by combining it with dexametha- spinal injury and focal or global ischaemia sone postoperatively for a few days."17 have shown a variable degree of protection Gammahydroxybutyrate drops blood pres- after treatment with the lazaroid U-74006F sure and its administration involves a consid- so that it is now undergoing large scale clini- erable sodium load."18 Propofol cal trials in head injury and subarachnoid (di-isopropylphenol) is widely used but care haemorrhage. Recently, the steroid compo- has to be taken to avoid hypotension. nent of the lazaroid molecule has been Propofol also has free radical scavenging replaced by the anti-oxidant ring structure of effects.' 9 The ideal hypnotic agent awaits vitamin E. U78517F has greater in vitro lipid development. In the patient with cardiovascu- anti-oxidant properties than U-74006F. It is lar instability, lignocaine (1 -5mg/kg iv) may interesting that the iron chelator desferriox- have a place in lowering ICP.'20 This dose is amine may be helpful in treating the coma of as effective as thiopentone (3mg/kg iv). cerebral malaria and experimental vasogenic Epilepsy has long been known to increase oedema. Early results in severe head injuries ICP and increase the risk of cerebral with the oxygen radical scavenger-poly- ischaemia as a result of a massive increase in Management ofraised intracranial pressure 855

cerebral electrical activity and oxidative at normal full-term. In the neonate, much J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from metabolism: both metabolic demand and lower CBF values may be tolerated for CPP are embarrassed. Seizures must be longer. Many of the pathologies differ from treated aggressively. Over the last decade, the adult including birth asphyxia, posthaem- however, a more focal phenomenon has been orrhagic ventricular dilatation, craniocerebral revealed: inappropriate hypermetabolism in disproportion and the many metabolic and small areas of the brain in association with infective encephalopathies. Hyperaemia plays local release or failed reuptake of excitotoxic a greater role as a cause of raised ICP in chil- amino acids such as glutamate.'2' 122 For dren after head injury than in adults.'28 129 The example, subdural haemorrhage in the rat is NIH Traumatic Data Bank of severe head accompanied by increased glutamate concen- injuries revealed that diffuse brain swelling trations in the with increased occurs twice as often in children (aged 16 local cerebral glucose utilisation and late years or younger) as in adults. A total of 53% ischaemic brain damage. An NMDA antago- of children with diffuse swelling died com- nist will protect against such damage'23 just as pared with a mortality rate of 16% in those NMDA antagonists reduce delayed neuronal without. The skull may expand in children loss and the volume of cerebral infarction where fusion of the sutures has not occurred. after middle cerebral artery occlusion, pro- Controlled trials of therapy in the various vided such agents are given prophylactically pathologies are made difficult by the very or in some cases within an hour of occlu- small numbers of patients seen in each cen- sion.'24 125 The therapeutic window of oppor- tre. As with adults, there is a wide diversity of tunity is very short unless it is surmised that opinion on the use of barbiturate coma, glutamate release occurs at various times after steroids and mannitol. Even fluid restriction a head injury, not just at the moment of still has its advocates in the neonates, and for impact. Such impressive experimental data inappropriate secretion of ADH. The dangers has led to Phase I and II trials of glutamate of fluid restriction, based on assuming that receptor antagonists in patients after severe SIADH is the common cause of hypona- head injury. The results are awaited with traemia after intracranial insults, have been keen interest. Ironically, that old-fashioned documented very clearly in studies of adults treatment for severe head injuries-rectal after SAH. magnesium chloride-is now known to be a noncompetitive NMDA receptor antagonist and reduces the infarct volume after middle Summary cerebral occlusion in the rat.'26 This review has been written at an unfortu- nate time. Novel questions are being asked of the old therapies and there is an abundance Targetted therapy of new strategies both to lower ICP and pro- It is clear that there are many causes of raised tect the brain against cerebral ischaemia. In ICP. The Edinburgh school has attempted7 the United Kingdom, the problem is to to define which therapy should be selected for ensure that appropriate patients continue to each cause of intracranial hypertension. They be referred to centres where clinical trials of suggest that hypnotic agents may be most high quality can be undertaken. One of the logically targetted at younger patients with success stories of the past decade has been http://jnnp.bmj.com/ diffuse congestive brain swelling with pre- the decline in the number of road accidents served cerebral electrical activity, jugular as a result of seat belt legislation, improve- venous oxygen saturation over 75%, a pulse- ments in car design and the drink/driving respiratory ratio in the ICP trace of over 2, laws. Hence, fortunately there are fewer preserved CBF CO2 reactivity and absence of patients with head injuries to treat and it is a diastolic notch on TCD recordings of MCA even more important that patients are appro-

velocity. Mannitol may be best used in priately referred if studies to assess efficacy of on October 1, 2021 by guest. Protected copyright. patients of any age with focal lesions, a low the new strategies are not to be thwarted. The cerebral perfusion pressure and reasonably nihilistic concept that intensive investigation preserved autoregulation. If arterial blood with ICP monitoring for patients with diffuse pressure is low despite colloid then inotropic head injury or brain swelling following evacu- agents such as Dobutamine or dopamine ation of a haematoma or a contusion has no should be used. SAH with raised ICP is best proven beneficial effect on outcome, requires managed by CSF drainage accepting that revision. A cocktail of therapies may be there is probably an increased risk of rebleed- required that can be created only when ing. patients are monitored in sufficient detail to reveal the mechanisms underlying their indi- vidual ICP problem. Children Ethical problems may arise over how The management of raised ICP in childhood aggressively therapy for intracranial hyperten- must take account of a number of factors.4 127 sion should be pursued and for how long. The critical values for ICP, ABP and CPP are There has always been the concern that cra- lower, the younger the child. The normal ICP nial decompression or prolonged barbiturate in the newborn is probably of the order of coma may preserve patients but with unac- 2-4 mmHg. ABP at birth is about 40 mmHg, ceptably severe disability. Some patients may 80/55 by one year and 90/60 during the early be salvaged from herniating with massive school years. CPP rises from 28 mmHg at cerebral infarction with the use of osmother- 28-32 weeks gestational age of 37-5 mmHg apy but is the outcome acceptable?"30 Similar 856 Pickard, Czosnyka

cephalus. Brain 1982;105:65-86. considerations apply to some children with J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from 27 Czosnyka M, Batorski L, Laniewski P, Maksymowicz W, metabolic encephalopathies. Where such con- Koszewski W, Zaworski W. A computer system for the siderations have been scrutinised in patients identification of the cerebrospinal compensatory model. Acta Neurochir (Wien) 1990;105:112-6. with severe head injury, the whole spectrum 28 Symon L, Dorsch NWC, Stephens RJ. Pressure waves in of outcomes appears to be shifted so that the so-called low-pressure hydrocephalus. Lancet 1972; 2:1291-2. number of severe disabilities and persistent 29 Pickard JD. Teasdale GM, Matheson M, Wyper DJ, et vegetative states are not increased. However, al. Intraventricular pressure waves-the best predictive test for shunting in normal pressure hydrocephalus. In: it is important to be sensitive to such issues Shulman K, et al, eds. Intracranial pressure IV. Berlin: based on experience of the particular cause of Springer-Verlag 1980:498-500. 30 Laniewski P, Czosnyka M, Maksymowicz W. raised intracranial pressure in a given age Continuous analysis of the intracranial pressure wave- group. form as a method of autoregulatory reserve assessment. ICP VIII. Berlin: Springer Verlag (in press). 31 Miller JD, Garibi J, Pickard JD. Induced changes in cere- We are indebted to Dr J M Turner, Mr E Guazzo and Mr P brospinal fluid volume. Effects during continuous Kirkpatrick for their comments on the manuscript monitoring of ventricular fluid pressure. Arch Neurol 1973;28:265-9. 32 Aaslid R, Markwalder T-M, Nornes H. Non-invasive 1 Jennett B, Teasdale G. Management of head injuries. transcranial Doppler ultrasound recording of flow Philadelphia: F A Davies. 1981. velocity in basal cerebral arteries. Jf Neurosurg 1982;57: 2 Miller JD, Jones PA, Dearden NM, Tocher JL. Progress 769-74. in the management of head injury. Br J Surg 1992; 33 Newell DW, Aaslid R. Transcranial doppler. New York: 79:60-64. Raven Press, 1992. 3 Pickard JD, Bailey S, Sanderson H, Rees M, 34 Martin JL, Perry S, Pickard JD. Cerebral blood flow and Garfield JS. Steps towards cost-benefit analysis of Doppler flow velocity: different responses to three regional neurosurgical care. BMJ 1990;301:629-35. vasodilators. Jf Cereb Blood Flow Metab 1991;11:suppl 4 Minns RA. Problems of intracranial pressure in child- 2:S455. hood. Clinics in developmental medicine 113/114. 35 Chan KH, Miller JD, Dearden NM, Andrews PJD, London: MacKeith Press, 1991:1-458. Midgley S. The effect of cerebral perfusion pressure 5 Langfitt TW. Increased intracranial pressure. Clinical upon changes in middle cerebral artery flow velocity Neurosurgery 1969;16:436-71. and jugular venous bulb oxygen saturation after severe 6 Miller JD. Normal and increased intracranial pressure. head injury. JfNeurosurg 1992;77:55-61. In: Miller JD, ed. Northfield's surgery of the central 36 Nelson RJ, Czosnyka M, Pickard JD, Maksymowicz W, 2nd ed, London: Blackwell, 1987:7-57. Perry S, Martin JL, Lovick AHJ. Experimental aspects 7 Miller JD, Dearden NM. Measurement, analysis and the of cerebrospinal hemodynamics: the relationship management of raised intracranial pressure. In: between blood flow velocity waveform and cerebral Teasdale GM, Miller JD, eds. Current neurosurgery autoregulation. Neurosurgery 1992;31:705-10. Edinburgh: Churchchill Livingstone, 1992:119-56. 37 Chan KH, Dearden NM, Miller JD. The significance of 8 Saul TG, Ducker TB. Effect of intracranial pressure posttraumatic increase in cerebral blood flow velocity: monitoring and aggressive treatment on mortality in a transcranial Doppler ultrasound study. Neurosurgery severe head injury. J Neurosurg 1982;56:498-503. 1992;30:697-700. 9 Welch K. The intracranial pressure in infants. J 38 Chan KH, Dearden NM, Miller JD, Midgley S, Piper Neurosurg 1980;52:693-9. IR. Transcranial Doppler waveform differences in 10 Lundberg N. Continuous recording and control of hyperemic and nonhyperemic patients after severe head ventricular fluid pressure in neurosurgical practice. injury. Surg Neurol 1992;38:433-6. Acta Psych Neurol Scand 1960;36,Suppl 149:1-193. 39 Chan KH, Miller JD, Dearden NM. Intracranial blood 11 Avezaat CJJ, von Eijndhoven JHM, Wyper DJ. flow velocity after head injury: relationship to severity pulse pressure and intracranial of injury, time, neurological status and outcome. Jf volume-pressure relationships. 7 Neurol Neurosurg Neurol Neurosurg Psychiatry 1992;55:787-91. Psychiatry 1979;42:687-700. 40 Schalen W, Messeter R, Nordstrom CH. Cerebral 12 Czosnyka M, Laniewski P, Batorski L, et al. Remarks on vasoreactivity and the prediction of outcome in severe amplitude-pressure characteristic phenomenon. In: traumatic brain lesions. Acta Anaesthesiol Scand Hoff JT, Betz AL, eds. Intracranial pressure VII. Berlin: 1991;35:113-22. Springer Verlag 1989:255-9. 41 Reid A, Marchbanks RJ, Martin R, Pickard JD, Bateman 13 Obrist WD, Langfitt TW, Jaggi JL, et al. Cerebral blood N, Brightwell R. Mean intracranial pressure monitor- flow and metabolism in comatose patients with acute ing by an audiological technique-a pilot study. J head injury. Relationship to intracranial hypertension. Neurol Neurosurg Psychiatry 1989;52:610-2. J Neurosurg 1984;61:241-55. 42 Campkin TV, Turner JM. Neurosurgical anaesthesia and 14 Bingham RM, Procaccio F, Prior PF, Hinds CJ. intensive care. 2nd ed. London: Butterworth, 1986.

Cerebral electrical activity influences the effects of eto- 43 Robertson CS, Narayan RH, Gokaslan ZL, et al. http://jnnp.bmj.com/ midate on cerebral perfusion pressure in traumatic Cerebral arteriovenous oxygen difference as an esti- coma. BritJAnaes 1985;57:843-8. mate of cerebral blood flow in comatose patients. J 15 Nordstrom CH, Messeter K, Sundbarg G, et al. Cerebral Neurosurg 1989;70:222-30. blood flow, vasoreactivity and oxygen consumption 44 Wyatt JS, Cope M, Delpy DT, Wray S, Reynolds EOR. during barbiturate therapy in severe traumatic brain Quantification of cerebral oxygenation and haemody- lesions. J Neurosurg 1988;68:424-3 1. namics in sick newborn infants by near infrared spec- 16 Rosner MJ, Becker DP. Origins and evolution of plateau troscopy. Lancet 1986;2:1063-6. waves. Experimental observations and theoretical 45 Jennett B, Teasdale G, Galbraith S, Pickard JD, et al. model. J Neurosurg 1984;60:312-24. Severe head injuries in three countries. Jf Neurol 17 Newell DW, Aaslid R, Stooss R, Reulen HJ. The rela- Neurosurg Psychiatry 1977;40:291-8. of flow fluctuations to intracra- 46 Miller Butterworth Gudeman et al. Further tionship blood velocity JD, JF, SK, on October 1, 2021 by guest. Protected copyright. nial pressure B waves. J Neurosurg 1992;76:415-21. experience in the management of severe head injury. J 18 Kirkham FJ, Levin SD, et al. Transcranial pulsed Neurosurg 1981;54:289-99. Doppler ultrasound findings in brainstem death. Jf 47 Ropper AH, O'Rourke D, Kennedy SK. Head position, Neurol Neurosurg Psychiatry 1987;50: 1504-1513. intracranial pressure and compliance. 1982; 19 Hassler W, Steinmetz H, Pirschel J. Transcranial 32:1288-91. Doppler study of intracranial circulatory arrest. J7 48 Rosner MJ, Coley IB. Cerebral perfusion pressure, ICP Neurosurg 1989;71:195-201. and head elevation. JNeurosurg 1986;65:636-41. 20 Marmarou A, Shulman K, Larmorgese J. 49 Feldman Z, Kanter MJ, Robertson CS, et al. Effect of Compartmental analysis of compliance and outflow head elevation on ICP, CPP and CBF in head injured resistance of the cerebrospinal fluid outflow system. Jf patients. J Neurosurg 1992;76:207-1 1. Neurosurg 1975;43:523-34. 50 Kanter MJ, Robertson CS, Sheinberg MA, et al. 21 Czosnyka M, Pickard JD, Whitehouse H, Piechnik S. Changes in cerebral hemodynamics with head elevated The hyperaemic response to a transient reduction in vs head flat. In: Avezaat C, ed. Intracranialpressure VIII cerebral perfusion pressure-a modelling study. Acta (In press). Neurochirurgica (Wien) 1992;115:90-7. 51 Hasan D, Vermeulen M, Wijdicks EFM, et al. Effect of 22 Marmarou A, Maset AL, Ward JD, et al. Contribution of fluid intake and hypertensive treatment on cerebral CSF and vascular factors to elevation of ICP in ischaemia after subarachnoid haemorrhage. severely head-injured patients. J Neurosurg 1987; 1989;20:1511-5. 66:883-90. 52 Nelson RJ, Roberts J, Rubin C, Walker V, Ackery DM, 23 Czosnyka M, Harris NG, Pickard JD. CO2 cerebrovas- Pickard JD. Association of hypovolaemia after SAH cular reactivity as a function of perfusion pressure-a with computer tomographic scan evidence of raised modelling study. Acta Neurochirurgica 1993;121: intracranial pressure. Neurosurgery 1991;29:178-82. 159-65. 53 Marmarou A, Anderson RL, Ward JD, et al. Impact of 24 Guillaume J, Janny P. Manometrie intracranienne con- ICP instability and hypotension on outcome in patients tinue. interet de la methode et premiers resultats. with severe head trauma. JNeurosurg 1991;75:S59-66. Revue Neurologique 195 1;84:131-42. 54 Muizelaar JP. Induced arterial hypertension in the treat- 25 Katzman R, Hussey F. A simple constant-infusion ment of high intracranial pressure. In: Hoff JT, Betz manometric test for measurement of CSF absorption. I AL, ed. Intracranial pressure VII. Springer-Verlag, Rationale and method. Neurology 1970;20:534-44. 1989:508-10. 26 Borgesen SE, Gjerris F. The predictive value of conduc- 55 Rosner MJ, Rosner SD. Cerebral perfusion pressure tance to outflow of CSF in normal pressure hydro- management of head injury. In: Avezaat C, ed. Management of raised intracranial pressure 857

Intracranial pressure VIII. (In press) furosemide combined with varying doses and adminis- 56 Shimoda M, Oda S, Tsugane R, Sato 0. Intracranial tration rates of mannitol. . Neurosurg 1987;66:440-6. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from complications of hypervolemic therapy in patients with 86 Smedema RJ, Gaab MR, Heissler HE. A comparison delayed ischemic deficit attributed to vasospasm. J study between mannitol and glycerol therapy in reduc- Neurosurg 1993;78:423-9. ing ICP. In: Avezaat, ed. Intracranial pressure VIII. (In 57 Fitch W, Pickard JD, Tamura A, Graham DI. Effects of press). hypotension induced with sodium nitroprusside on the 87 Muizelaar JP, van der Poel HG, Li Z, Kontos HA, cerebral circulation before, and one week after, the Levasseur JE. Pial arteriolar vessel diameter and CO2 subarachnoid injection of blood. .7 Neurol Neurosurg reactivity during prolonged hyperventilation in the rab- Psychiatry 1988;51:88-93. bit. J Neurosurg 1988;69:923-7. 58 Nelson PB, Seif SM, Maroon JC, Robinson AG. 88 Ward JD, Choi S, Marmarou A, et al. Effect of prophy- Hyponatraemia in intracranial disease: perhaps not the lactic hyperventilation on outcome in patients with syndrome of inappropriate secretion of antidiuretic severe head injury. ICP VII. In: Hoff IT, Betz AL, eds. hormone (SIADH). _7 Neurosurg 198 1;55:938-4 1. Berlin: Springer-Verlag, 1989:630-4. 59 Nelson RJ, Perry S, Burns ACR, Roberts J, Pickard JD. 89 Enevoldsen EM, Cold G, Jensen FT, et al. Dynamic The effects of hyponatraemia and subarachnoid haem- changes in regional CBF, intraventricular pressure, orrhage on the cerebral vasomotor responses of the CSF pH and lactate levels during the acute phase of rabbit. .7 Cereb Blood Flow Metab 199 1;11:661-6. head injury. J Neurosurg 1976;44:191-214. 60 Myers RE. Anoxic brain pathology and blood glucose. 90 De Salles AAF, Kontos HA, Becker DP, et al. Prognostic Neurology 1976;34:345. significance of ventricular CSF lactic acidosis in severe 61 Marie C, Bralet J. Blood glucose level and morphological head injury. _7 Neurosurg 1986;65:615-24. brain damage following cerebral ischaemia. 91 Akiota T, Ota K, Matsumato A, et al. The effect of Cerebrovascular and Brain Metab Reviews 1991 ;3:29-38. THAM on acute intracranial hypertension. An experi- 62 Reed DJ, Woodbury DM. Effect of hypertonic urea on mental and clinical study. In: Beks JWF, Bosch DA, cerebrospinal fluid pressure and brain volume. 7 Physiol Brock M, eds. Intracranial pressure III. Berlin: Springer- (Lond) 1963;164:252-64. Verlag, 1976:219-33. 63 Pappius HM, Dayes LA. Hypertonic urea-its effect on 92 Gordan E, Rossanda M. Further studies in cerebrospinal the distribution of water and electrolytes in normal and fluid acid-base status in patients with brain lesions. edematous brain tissue. Arch Neurol 1965;13:395-402. Acta Anaesth Scand 1970;14:97-109. 64 Ravassin P, Abou-Madi M, Archer D, Chiolero R, 93 Gaab MR, Seegers K, Goetz C. THAM (tromethamine, Freeman J, Trop D, de Tribolet N. Changes in CSF "tris-buffer"): effective therapy of traumatic brain pressure after mannitol in patients with and without swelling. In: HoffJT, Betz AL, eds. Intracranial pressure elevated CSF pressure. _7 Neurosurg 1988;69:869-76. VII. Berlin: Springer-Verlag, 1989:616-9. 65 James HE, Harbaugh RD, Marshall LF, Shapiro HM. 94 Wolf AL, Levi L, Marmarou A, et al. Effect of THAM The response to multiple therapeutics in experimental upon outcome in severe head injury: a randomized vasogenic . In: Shulman K, Marmarou A, Miller prospective clinical trial. _7 Neurosurg 1993;78:54-9. JD, et al, eds. ICP IV. Berlin, Springer-Verlag. 1980; 95 Pickard JD, Mackenzie ET. Inhibition of prostaglandin 272-6. synthesis and the response of baboon circulation to 66 Rosenberg GA. Brain fluids and metabolism. Oxford: carbon dioxide. Nature New Biology 1973;245:187-8. Oxford University Press, 1990. 96 Edwards AD, Wyatt JS, Richardson C, et al. Effects of 67 Inao S, Kuchiwaki H, Wachi A, et al. Effect of mannitol indomethacin on cerebral haemodynamics in very on intracranial pressure-volume status and cerebral preterm infants. Lancet 1990;335:1491-5. haemodynamics in brain oedema. Acta Neurochirurgica 97 Jensen K, Ohrstrom J, Cold GE, Astrup J. The effects of 1990;Suppl5l :401-3. indomethacin on intracranial pressure, cerebral blood 68 Bell BA, Smith MA, Kean DM, et al. Brain water mea- flow and cerebral metabolism in patients with severe sured by magnetic resonance imaging: correlation with head injury and intracranial hypertension. Acta direct estimation and change following mannitol and Neurochir (Wien) 1991;108:116-21. dexamethasone. Lancet 1987;1:66-9. 98 Sullivan LP, Stears JC, Ringer SP. Resolution of 69 Kaufmann AM, Cardoso ER. Delayed cerebral accumu- syringomyelia and chiari malformation by ventriculo- lation of mannitol in vasogenic edema. In: Avezaat C, atrial shunting in a patient with pseudotumour cerebri ed. Intracranial pressure VIII. (In press). and a lumbo-peritoneal shunt. Neurosurgery 1988;22: 70 Burke AM, Quest DO, Chien S, et al. The effects of 744-7. mannitol on blood viscosity. . Neurosurg 1981;55: 99 Editorial. Cranial decompression. Lancet 1988; 1: 1204. 550-3. 100 Gower DJ, Lee KS, McWhorter JM. Role of subtempo- 71 Miller JD, Leech P. Effects of mannitol and steroid ral decompression in severe closed head injury. therapy in intracranial volume pressure relationships..7 Neurosurgery 1988;23:417-22. Neurosurg 1975;42:274-81. 101 Umezewa H, Shima K, Chigasaki H, Ishii S. Effects of 72 Muizelaar JP, Wei EP, Kontos HA, Becker DP. the pressure gradient on hydrostatic brain edema. In: Mannitol causes compensatory cerebral vasoconstric- Hoff JT, Betz AL, eds. Intracranial pressure VII. Berlin: tion and vasodilatation in response to blood viscosity Springer-Verlag, 1989:953-6. changes. JNeurosurg 1983;59:822-8. 102 Rittierodt M, Gaab MR. Traumatic brain swelling and 73 Auer LM, Haselsberger K. Effect of intravenous manni- operative decompression: a prospective investigation. tol on cat pial arteries and veins during normal and In: Avezaat C, et al, ed. Intracranial pressure VIII. elevated intracranial pressure. Neurosurgery 1987; Berlin: Springer-Verlag, (In press). http://jnnp.bmj.com/ 21:142-6. 103 Kirkham FJ, Neville BGR. Successful management of 74 Harrison MJG. Influence of haemocrit in the cerebral severe intracranial hypertension by surgical decompres- circulation. Cerebrovascular and Brain Metab Reviews sion. Developmental Medicine Child Neurol 1986;28: 1989;1:55-67. 506-9. 75 Brown MM, Wade JPH, Marshall J. Fundamental 104 Cooper PR, Moody S, Clark WK, et al. Dexamethasone importance of arterial oxygen content in the regulation and severe head injury: a prospective double blind trial. of cerebral blood flow in man. Brain 1985;108:81-93. . Neurosurg 1979;51:307-16. 76 Purves MJ. The Physiology of the cerebral circulation. 105 Saul TF, Ducker TB, Salomon M, et al. Steroids in Cambridge: Cambridge University Press, 1972. severe head injury. A prospective randomized clinical 77 Bruce DA, Langfitt TW, Miller JD, et al. Regional cere- trial. _7 Neurosurg 1981;54:596-600. bral blood flow, intracranial pressure, and brain metab- 106 Braakman R, Schouten HJD, Dishoeck BMV, olism in comatose patients. _7 Neurosurg 1973;38: Minderhoud JM. Megadose steroids in severe head on October 1, 2021 by guest. Protected copyright. 131-44. injury: results of a prospective double blind clinical 78 Johnston IH, Harper AM. The effect of mannitol on trial. J Neurosurg 1983;58:326-30. cerebral blood flow. An experimental study. .7 107 Giannotta SL, Weiss MH, Apuzzo MLJ, Martin E. High Neurosurg 1973;38:461-71. dose glucocorticoids in the management of severe head 79 Mendelow AD, Teasdale GM, Russell T, et al. Effect of injury. Neurosurgery 1984;15:497-501. mannitol on cerebral blood flow and cerebral perfusion 108 Dearden NM, Gibson JB, McDowall DG, et al. Effect of pressure in human head injury. _7 Neurosurg 1985; high dose dexamethasone on outcome from severe 63:43-8. head injury. .7 Neurosurg 1986;64:81-8. 80 Jafar JJ, Johns LM, Mullan SF. The effect of mannitol on 109 Floyd RA, Camey JM. Protection against oxidative dam- cerebral blood flow..7 Neurosurg 1986;64:754-9. age to CNS by a-Phenyl-tert-butyl Nitrone and other 81 Muizelaar JP, Lutz HA, Becker DP. Effect of mannitol spin-trapping agents: a novel series of nonlipid free on ICP and CBF and correlation with pressure radical scavengers. In: Marangos PJ, Lal H, eds. autoregulation in severely head-injured patients. _7 Emerging strategies in neuroprotection, Boston: Birkauser, Neurosurg 1984;61:700-6. 1992:252-72. 82 Meyer FB, Anderson RE, Sundt TM, Yaksh TL. 110 Halliwell B, Gutteridge JMC. Free radicals in biology and Treatment of experimental focal cerebral ischaemia medicine. Oxford: Clarendon Press, 1985. with mannitol. Assessment by intracellular pH, cortical 111 Hall ED. Lazaroids: novel cerebroprotective antioxi- blood flow and EEG. JTNeurosurg 1987;66:109-15. dants. In: Marangos PJ, Lal H. Emerging strategies in 83 Takagi H, Saito T, Kitahara T, Morii SM, Ohwada J, neuroprotection. Boston: Birkauser, 1992:224-37. Yada K. The mechanism of the ICP reducing effect of 112 Muizelaar JP, Marmarmou A, Young HA, et al. mannitol. In: Ishii S, Nagai H, Brock M, eds. ICP V. Improving the outcome of severe head injury with the Berlin: Springer-Verlag, 1983:729-33. oxygen radical scavenger polyethylene glycol-conju- 84 Smith HP, Kelly DL, McWhorter JM, et al. Comparison gated superoxide dismutase: a Phase II trial. . of mannitol regimes in patients with severe head injury Neurosurg 1993;78:375-82. undergoing ICP monitoring. .7 Neurosurg 1986;65: 113 Welsh FA, Sims RE, Harris VA. Mild hypothermia pre- 820-4. vents ischemic injury in gerbil hippocampus. .7 Cereb 85 Roberts PA, Pollay M, Engles C, Pendleton B, Reynolds Blood FHow Metab 1990;10:557-63. E, Stevens FA. Effect on intracranial pressure of 114 Schwartz M, Tator CH, Rowed DW, et al. The 858 Pickard, Czosnyka

University of Toronto Head Injury Treatment Study: a bolism after acute subdural haematoma is ameliorated prospective randomized comparison of pentobarbitat by a competitive NMDA antagonist. Neurotrauma and mannitol. Canad JNeurol Sci 1984;11:434-40. 1992;9:75-84. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.56.8.845 on 1 August 1993. Downloaded from 115 Ward JD, Becker DP, Miller JD, et al. Failure ofprophy- 124 Gill R, Foster AC, Woodruff GN. MK801 is neuropro- lactic barbiturate coma in the treatment of severe head tective in gerbils when administered during the post- injury. JNeurosurg 1985;62:383-8. ischaemic period. Neuroscience 1988;25:847-56. 116 Eisenberg HM, Frankowski RF, Condant CG, et al. 125 Park CK, Nehls DG, Graham DI, Teasdale GM, High dose barbiturate control of elevated intracranial McCulloch J. The glutamate antagonist MK801 pressure in patients with severe head injury. reduces focal ischaemic brain damage in the rat. Ann Neurosurg 1988;69: 15-23. Neurol 1988;24:543-55. 117 Batjer HH. Cerebral protective effects of Etomidate: 126 Izumi Y, Roussel S, Pinard E, Seylaz J. Reduction of experimental and clinical aspects. Cerebrovasc Brain infarct volume by magnesium after middle cerebral Metab Rev 1993;5:17-32. artery occlusion in rats. Cereb Blood Flow Metab 118 Leggate JRS, Dearden NM, Miller JD. The effects of 1991;11:1025-30. gammahydroxybutyrate and thiopentone on ICP in 127 Levene M. Neonatal neurology. Edinburgh, Churchill severe head injury. In: Miller JD, Teasdale GM, et al Livingstone. 1987. eds. Intracranial pressure VI. Berlin: Springer-Verlag, 128 Bruce DA, Alavi A, Bilaniuk, et al. Diffuse cerebral 1986:754-9. swelling following head injuries in children: the syn- 119 Murphy PG, Myers DS, Davies MJ, et al. The antioxi- drome of "malignant brain edema". Neurosurg dant potential of propafol (2,6-diisopropylphenol). Brit 1981;54:170-8. JAnaes 1992;68:613-8. 129 Aldrich EF, Eisenberg HM, Saydjari C, et al. Diffuse 120 Bedford RF, Persing JA, Pobereskin L, Butler A. brain swelling in severely head-injured children. Lidocaine or thiopental for rapid control of intracranial Neurosurg 1992;76:450-4. hypertension? Anaesth Anaig 1980;59:435-7. 130 Harrison MJG. Is shrinking the brain a good thing after 121 Meldrum B. Protection against ischaemic neuronal dam- cerebral infarction? In: Warlow C, Garfield J, eds. age by drugs acting on excitatory neurotransmission. Dilemmas in the management of the neurological patient. Cerebrovascular Brain Metab Reviews 1990;2:27-57. Edinburgh: Churchill Livingstone, 1984:62-73. 122 Choi DW. Methods of antagonizing glutamate neurotox- 131 Pullen RGL, De Pasquale M, Cserr HF. Bulk flow of icity. Cerebrovasc Brain Metab Rev 1990;2:105-47. cerebrospinal fluid into brain in response to acute 123 Inglis FM, Kuroda Y, Bullock R. Glucose hypermeta- hyperosmolality. AmerJ Physiol 1987;253:538-45. http://jnnp.bmj.com/ on October 1, 2021 by guest. Protected copyright.