The cerebral vessels change diameter inversely with changing perfusion pressure: as CPP rises, the vessels constrict and as CPP falls the vessels dilate, such that blood flow is kept constant over a wide range of CPP (Figure 45.2a). This pressure autoregulation is thought to be controlled by local myogenic responses of the vessel wall to changes in intra-arterial pressure. At pressures above and below this range of 6.7–20 kPa (50–150 mmHg), cerebral perfusion becomes pressure-passive and increases or decreases in direct proportion to changes in CPP. The autoregulatory range varies with age, being shifted to the left in newborns and to the right in those with chronic hypertension. The latter is important to remember to avoid overtreating systolic blood pressure in such patients and thus incur the risk of cerebral ischaemia at the lower limits of autoregulation. Alternatively, cerebral perfusion above normal can be caused by acute hypertension overcoming the upper limits of autoregulation. This may lead to cerebral oedema secondary to increased hydrostatic pressures (hypertensive encephalopathy) and potentially lead to seizures or cerebral haemorrhage.

Figure 45.2 (a) The relationship between the partial pressure of oxygen (PO₂) and carbon dioxide (PCO₂) and cerebral blood flow (CBF). (b) The relationship between mean arterial pressure (MAP) and cerebral blood flow under normal circumstances, illustrating the range of autoregulation.

PaO₂ AND PaCO₂ EFFECTS

A second group of factors control CBF through an influence on the local metabolic milieu. Prominent in this mechanism are oxygen and carbon dioxide. Arterial content or partial pressure of oxygen in the normal or hyperoxic ranges causes very little change in CBF. Perhaps this represents a demand for another nutrient (i.e. glucose) or a need to remove waste products (i.e. carbon dioxide or metabolic acid). With the onset of hypoxaemia (PaO₂ 60 mmHg or 8 kPa), there is a prompt increase in CBF proportional to the decrease in blood oxygen content, in order to maintain oxygen delivery constant (Figure 45.2b).

There is also a direct relationship between CBF and PaCO₂, such that cerebral perfusion increases with increasing PaCO₂ (Figure 45.2b). This probably represents the need of the brain to maintain homeostatic pH by removing metabolic breakdown products more efficiently by increased blood flow. Unlike the response to oxygen, the CBF response to changes in PaCO₂ is dramatic in the physiological range, such that for every 0.13 kPa (1.0 mmHg) change in PaCO₂ there is a 1–2 ml/min per 100 g tissue change in CBF. Therefore, an increase in PaCO₂ to 10.6 kPa (80 mmHg) will increase CBF to approximately 100 ml/min per 100 g and a decrease in PaCO₂ to 2.7 kPa (20 mmHg) will decrease CBF to 25 ml/min per 100 g. Thus:

• doubling PaCO₂ doubles CBF

• halving PaCO₂ halves CBF within this range

Understanding this basic physiology will make treatment logical (see below), as increases in CBF often lead to increases in cerebral blood volume, which in turn can increase ICP – a common cause of cerebral ischaemia.

CEREBRAL INJURY

Vascular insufficiency or disruption, trauma, tumour, infection, inflammation and metabolic and nutritional derangement can all cause damage. Whatever the cause, the mechanism of injury is usually hypoxic and/or ischaemic injury.

• Hypoxia is lack of oxygen.

• Ischaemia is lack of flow.

Therefore, it is possible to have high flow but no oxygen or, alternatively, no flow but plenty of oxygen. At the cellular level the consequence is the same.

The protective mechanisms afforded by the ability to increase CBF manifold in response to hypoxaemia and the generous oversupply of nutrients under normal conditions allow for sufficient blood flow, with maintenance of oxygen delivery and supply of other nutrients in many cases. Limitation in the ability to alter CBF, whether through cerebrovascular disease restricting perfusion or intracranial masses, oedema and increased ICP exerting excess local pressure, and altering blood flow, will cause hypoxic/ischaemic injury. Systemic problems such as severe or prolonged hypoxaemia will eventually disturb systemic circulatory homeostasis, leading to hypotension and eventual ischaemia. Thus hypoxic and ischaemic injury may be considered synonymous, although there may be aetiological differences between them. A more important distinction is to be made between global and focal hypoxic or ischaemic insults.

GLOBAL HYPOXIC/ISCHAEMIC INSULTS

These types of insult are caused by hypoxaemia and cardiovascular insufficiency or arrest, respectively. These are usually sudden, short and severe. If there is to be recovery, prompt return of oxygen delivery and spontaneous circulation are necessary. The recovery may be variable, depending on the severity and duration of the insult and the selective vulnerability of the cell types involved. Different mechanisms are responsible for reversible loss of cellular function and for irreversible cell death and there are also differences between the mechanisms that cause death of neurones, glia and endothelial cells. After 4–6 minutes of complete global ischaemia, there are signs of permanent histological damage in selective neuronal populations and the beginnings of neurological deficits in survivors. Outcome worsens significantly after 15 minutes of global ischaemia.¹

FOCAL HYPOXIC/ISCHAEMIC INSULTS

These insults often occur suddenly but are usually of more prolonged duration. Only a subtotal of the brain is affected and the injury may be less severe if the surrounding brain is preserved by collateral blood supply.

The area of the focal ischaemia supplied by end-arteries will result in cell death unless reperfusion is established rapidly. The periphery of an infarcted area is the ischaemic penumbra. Here, CBF is greater than at the infarcted core, but less than in the normal tissue around it. Animal studies suggest that the time course for infarction and irreversible damage to brain from a focal hypoxic/ischaemic event is around 30–60 minutes. The focus of therapeutic intervention is the penumbral area, on the basis that if blood flow can be normalised in this region, or pharmacological agents can be delivered despite the reduced blood flow, there is a potential for recovery. Conversely, failure to maintain this area will result in a coalescing of the penumbra into the infarcted area, as the ischaemic stimulus continues.

It is likely that the ongoing ischaemic penumbral area of a focal insult and transient whole-brain ischaemia (e.g. during early cardiac arrest, low-flow cardiopulmonary resuscitation (CPR) or elevated ICP states) are subject to similar pathophysiological processes. As pretreatment is usually impossible, prevention and treatment of the secondary insults are the focus in neuroprotection and tissue salvage strategies.

BRAIN ISCHAEMIC AND INFARCTION PROCESSES

Changes in normal physiology begin to occur when blood flow is reduced.

• At CBF below 50 ml/min per 100 g neurological function is impaired and there is slowing of the electroencephalogram (EEG).

• A CBF of 15–25 ml/min per 100 g results in loss of electrical activity.

• A CBF between 10 and 15 ml/min per 100 g can maintain adenosine triphosphate (ATP) levels sufficiently to support ionic pump function for a time, despite the lack of electrical activity and normal neurological function.

• At a CBF of 10 ml/min per 100 g membrane failure occurs, due to a critical loss of ATP, which causes ionic imbalance between the cell and the extracellular milieu. If prolonged or worsened, CBF at this level will lead to permanent neurological impairment as a result of cell death.

EFFECTS OF ISCHAEMIA

Ischaemia results in reduced available oxygen and glucose to support aerobic production of ATP. Levels of ATP are depleted within 2–3 minutes of complete ischaemia (animal studies). There is little brain storage of either glucose or oxygen, and ATP production during ischaemia relies on anaerobic glycolysis for as long as stores last. This results in continued ATP use, but suboptimal production of ATP to fuel aerobic metabolism, so a lactic acidosis develops. Loss of ATP causes failure of membrane ionic pump function, leading to an efflux of potassium and an influx of sodium, calcium and chloride ions, the beginnings of cytotoxic oedema.

This begins a cascade of events resulting in eventual cell death:

• Potassium leakage causes cell depolarisation, ion channel opening and the release of excitatory amino acid (EAA) neurotransmitters (glutamate and aspartate and others).

• EAA cause further depolarisation of neighbouring cells and distal, otherwise unaffected, parts of the brain which result in:

• sodium and chloride influx via kainate (K) and quisqualate (Q) receptors (with water worsening intracellular oedema)

• calcium influx via N-methyl-D-aspartate (NMDA) receptors

• release of excitatory transmitters

• conversion of phosphorylases, uncoupling of oxidative phosphorylation, activation of proteases, degradation of cystosolic protein and stimulation of lipases. Lipases liberate arachidonic acid and other free fatty acids that cause tissue damage via production of oxygen radicals and prostaglandins

Other effects within the cell influence DNA and RNA production, hence inhibiting protein production. This may explain why cellular and clinical recovery is partial, even with restoration of ionic equilibrium and near-normal ATP levels after successful reperfusion. Necrosis is thought to occur in the core of the cerebral infarct following acute vascular occlusion, with further neurodegeneration occurring more slowly in the penumbra, by apoptosis or release of various immunological mediators.²

EXTRACELLULAR EFFECTS

Leukocytes are thought to be major contributors to reperfusion injury in that:

• they may plug up small capillaries under conditions of low blood flow and prevent reflow in certain areas, thus hindering restoration of perfusion

• they may enhance production of oxygen radicals and begin a cascade of inflammatory mediators, which may potentiate cell destruction in injured tissue

This results in tissue oedema, which may affect the core lesion by narrowing blood vessels and worsening chances of reperfusion. Local mechanical compression of tissue or its blood supply can cause deterioration in neighbouring tissue. The alteration in volume is of extreme importance in the adult brain because the cranial vault is non-distensible, preventing accommodation to an expanding lesion. This will impede CBF both locally and globally.

Hypoxic/ischaemic insults are rarely predictable, so that the process is usually well established when clinicians intervene. It is therefore highly likely that the injury will be only partially remediable.

MANAGEMENT

In neurological injury, the initial aims are to provide basic support. Assessment of the airway and respiration are the first priority, closely followed by optimisation of the circulation.

In cases of head trauma, it is very important to prevent secondary brain injury. Reviews of intensive care practice have resulted in recommendations for treatment in this group of patients.³ These involve:

• institution of monitoring

• early treatment of hypotension, hypoxia, hyperthermia and intracranial hypertension

• maintenance of CPP

Following stroke, it is also important to optimise homeostasis and address any hypertension, hyperglycaemia, hyperthermia and intracranial hypertension, as these are independent factors of a poor prognosis.

HYPERTENSION

Most patients have elevated blood pressure in the early phase after an acute ischaemic stroke. The mechanism and effects of this blood pressure elevation are not well understood and the benefits of intervention remain debated. Most experts would not advocate antihypertensive treatment in the acute phase after stroke, unless the blood pressure is particularly high. Currently hypertension (systolic blood pressure > 180 mmHg) should be treated if thrombolytics are to be administered, based on the trials involving these agents in which the blood pressure was lowered prior to treatment.⁴ Ongoing studies such as Blood Pressure in Acute Stroke Collaboration (BASC) should help define optimal management.

HYPERGLYCAEMIA

Hyperglycaemia has also been associated with an increased mortality and reduced functional outcome after stroke irrespective of whether the patient has a diabetic history or not.⁵ Proposed explanations for this are:

• Hyperglycaemia may be directly toxic to the ischaemic brain.

• Decreased levels and reduced peripheral uptake of insulin in these patients result in increased cerebral glucose availability which may impair endothethelial-dependent vasodilatation.

• Hyperglycaemia may indicate dysglycaemia, which is associated with vascular disease and endothelial dysfunction, both of which may exacerbate ischaemic damage at the time of stroke.

Hyperglycaemia may result in a hypercoaguable state, with decreased plasma fibrinolytic activity, which may delay reperfusion and increase haemorrhagic infarct conversion in tissue plasminogen activator-treated patients.⁶

Insulin, as well as having a glucose-lowering effect, has also been shown to be directly neuroprotective and this is the subject of a number of ongoing trials ⁷ (Treatment of Hyperglycaemia in Ischaemic Stroke (THIS) trial, the Glucose Regulation in Acute Stroke Patients (GRASP) trial and the Glucose Insulin Stroke Trial UK (GIST-UK).

HYPERTHERMIA

An increased temperature increases cerebral metabolism, oxygen requirements, CBF and ICP. Hyperpyrexia following acute stroke adversely influences stroke severity, infarct size, functional outcome and mortality.⁸ A raised temperature should, therefore, be treated aggressively and any evidence of infection identified early and treated with appropriate antibiotics.

REVASCULARISATION

For global ischaemia following cardiac arrest, the theoretical principles are to re-establish systemic blood pressure and to consider the no-reflow phenomenon in certain areas of the brain. Current research is addressing attempts to increase blood pressure during CPR and to open up capillaries that may have collapsed during the arrest period.

For focal cerebral ischaemia:

• Start treatment as early as possible.

• Increase local CBF (fibrinolysis).

• Try to ameliorate ischaemic cascade (neuroprotection).

Ischaemic neuronal degeneration follows the onset of perfusion failure so that interventional therapy must begin immediately. Diagnostic imaging is required before using thrombolytic agents. For acute ischaemic stroke the National Institute of Neurological Disorders and Stroke (NINDS study) suggested efficacy if intravenous thrombolytic therapy is initiated within the first 3 hours of stroke symptoms (using recombinant tissue plasminogen activator).⁹

Current issues in implementing this treatment include:

• diagnosis, including imaging within this time frame

• the potential risks of thrombolysis

• accurately selecting patients eligible for thrombolysis

HAEMODILUTION

Current practice includes the use of hypervolaemic haemodilution and deliberate hypertension with augmentation of cardiac output using vasopressors and inotropic agents, particularly in the treatment of delayed ischaemic deficits after subarachnoid haemorrhage related to vasospasm. This therapy can be administered safely with intensive monitoring, but benefits are anecdotal, as it has not been tested with concurrent controls.¹⁰

Decreasing haematocrit and viscosity by haemodilution may improve flow and hence blood and oxygen delivery to areas with impaired arterial supply. Benefit in animal models of ischaemia has not been found in clinical trials of stroke. Trials of normovolaemic ¹¹ and hypervolaemic haemodilution¹² have had complications from volume therapy in patients with associated heart disease.

Theoretical neuroprotective benefits of human haemoglobin-derived blood substitutes have not translated into clinical benefit, largely due to adverse events, and therefore clinical development of substitutes were discontinued.¹³

INTRACRANIAL PRESSURE

Raised ICP can cause global ischaemia. Treating ICP requires knowledge of the three compartments contributing to ICP within a fixed cranium:

1 blood

2 relatively non-compliant brain tissue

3 cerebrospinal fluid (CSF)

The relationship between ICP and intracranial volume is described by a non-linear pressure–volume curve. At lower intracranial volumes the ICP remains low and reasonably constant. Any increases in intracranial volume are compensated for by decreases in intracerebral blood or CSF volume. If greater intracranial volumes persist, this compensation is lost and ICP rises considerably, despite relatively small increases in intracerebral blood volume. Eventually at high levels of ICP cerebrovascular responses are lost, ICP equals MAP and CPP is very low.¹⁴

To maintain adequate cerebral perfusion, treatment should be targeted at ensuring an adequate perfusion pressure and a reduction in ICP. MAP should be raised to a level at or above the usual pressure for that patient, within the zone of pressure autoregulation – hence knowledge of premorbid blood pressure is important. If the majority of the vasculature is autoregulating, raising the blood pressure may decrease vascular diameter and reduce blood volume within the cranium. If the cause of the raised ICP cannot be corrected (e.g. blood clot or brain tumour), then the focus should be to prevent secondary injury around the lesion. An increased ICP can cause further ischaemia, and a reduction in ICP may facilitate adequate perfusion to areas at risk. Whenever possible, the offending compartment should be treated primarily (e.g. tumour removal, blood evacuation, drainage of hydrocephalus and, on occasion, craniectomy). If this is not advisable, reducing the relative volumes of other compartments may improve compliance overall and reduce the ICP.

MONITORING OF ICP (Figure 45.3)

ICP-monitoring devices have been ranked by the Brain Trauma Foundation ¹⁴ on their accuracy, stability and ability to drain CSF. Intraventricular and intraparenchymal catheters are seen as most favourable, followed by subdural, subarachnoid and epidural devices.

Figure 45.3 Intracranial pressure (ICP) waveforms. A-waves are plateau waves of 50–100 mmHg, sustained for 5–15 minutes, associated with raised ICP and compromised cerebral blood flow. B-waves are small changes in pressure every 0.5–2 minutes, often associated with breathing patterns and possibly due to local variations in the partial pressure of oxygen and carbon dioxide. C-waves are low-amplitude oscillations with a frequency of about 5 per minute, associated with variation in vasomotor tone. Resps, respirations; BP, blood pressure.

INTRAVENTRICULAR DEVICES

ICP is measured from an external pressure transducer attached to an intraventricular catheter inserted in one of the lateral ventricles using the external auditory meatus as the reference point. This device allows CSF sampling for culture or therapeutic drainage, the administration of antibotics and the ability to rezero following insertion. The main disadvantage is the high infection rate associated with their use.^14,¹⁵ Infection rates increase significantly after 5 days of catheter insertion and with repeated use of the device’s flushing system. The use of antibiotic-impregnated catheters may lower the infection rate. Air, blood, tissue debris and ventricular collapse can all adversely affect measurement.

INTRAPARENCHYMAL DEVICES

ICP can be measured by either a miniature strain gauge pressure transducer mounted at the end of a thin catheter or by a fibreoptic cable which directs light to a miniature mirror at the end of the cable placed in brain tissue. These monitors are advantageous when lateral ventricles have collapsed secondary to raised ICP. This system is not dependent on fluid coupling for pressure transduction so waveform damping and artefacts are avoided. The infection rates are very low. Its main disadvantages include lack of recalibration in vivo, a tendency for baseline drift requiring transducer replacement, the inability to drain CSF and the ability to measure only localised pressure which may not be representative of the global ICP.¹⁵

Alternative invasive ICP monitoring includes epidural, subarachnoid and subdural devices, although these are not as accurate. All invasive ICP monitoring devices are associated with a risk of parenchymal damage with the potential for parenchymal and subdural bleeding. When it is not possible to insert an ICP-monitoring device, e.g. because of severe coagulopathy, ICP can be estimated from non-invasive alternatives, e.g. transcranial Doppler examination. However all ICP-monitoring devices have their limitations. Clarification should be sought by computed tomographic (CT) scanning if there is doubt about whether to intervene with treatment of an elevated ICP.

REGULATION OF INTRACRANIAL PRESSURE

HYPOTHERMIA

Injury to the central nervous system is temperature-dependent. Fever can make an existing neurological dysfunction more apparent and may worsen an ongoing dysfunction. Potential mechanisms by which hyperthermia worsens cerebral ischaemia may include:

• neurotransmitter and oxygen free radical production

• blood–brain barrier failure

• damaging depolarisations in the ischaemic penumbra

• impaired recovery of energy metabolism

• cytoskeletal proteolysis ²³

Hyperthermia is harmful so treatment of fever should be aggressive, using cooling blankets, cool water, cool intravenous fluids, fans and antipyretic medications.

Hypothermia has been known to offer protection for years. In particular, drowning victims who were hypothermic have survived long periods of ischaemia. The mechanism of ICP reduction is unknown but may be due to a reduction in intracranial blood volume secondary to cerebral vasoconstriction or to alterations in metabolism. The protective effect of cooling appears to be much greater than that explained by changes in metabolism alone.

In contrast to preventing hyperthermia, the use of induced hypothermia is more complex. It has been used for protection extensively in coronary artery bypass surgery where hypothermia (28–30°C) is commonplace and deep hypothermia (< 20°C) has allowed prolonged circulatory arrest for surgery such as high thoracic and giant cerebral aneurysms. Animal work suggests that moderate, systemic hypothermia reduces the cerebral oedema and death after injury to the cerebral cortex. In several centres cooling has been used as therapy for severe head injuries. Unfortunately, a study which evaluated the efficacy of hypothermia in head injuries was halted after the enrolment of 392 patients because the treatment was ineffective.²⁴ Cooling patients to 33°C within 8 hours after injury and maintaining hypothermia for 48 hours were not effective in improving the clinical outcome at 6 months and patients older than 45 years of age had a poorer outcome. This contrasted with two earlier studies,^25,²⁶ both of which demonstrated an improvement in outcome with cooling to 32°C.

In conclusion, at present:

• There is no evidence that hypothermic therapy after traumatic brain injury reduces risk of death or neurological outcome.

• Hypothermia may put the patient at more harm, with the risk of pneumonia, arrhythmias and coagulopathies.

• It may be more problematic in patients over 45 years of age.

In stroke patients there are no randomised controlled trials of hypothermia. Animal models suggest that hypothermia reduces ischaemic stroke lesion size.²⁷ A small uncontrolled study of moderate hypothermia found reduced mortality from an expected value of 78% to 44% in patients with severe middle cerebral artery infarction.²⁸ There are ongoing studies such as the Nordic Cooling Stroke Study (NOCSS).

In cardiac arrest the situation is changing. Although 50 years ago therapeutic hypothermia was used post cardiac arrest, it was abandoned because of the uncertain benefit. In the 1980s, animal studies indicated benefit with mild (32–35°C) rather than moderate or deep hypothermia. Two randomised clinical trials of hypothermia, after witnessed out-of-hospital cardiac arrest, demonstrated improved survival and neurological outcome after 12 or 24 hours of hypothermia at 33°C.^29,³⁰ The evidence from these two relatively small trials has led to recommendations as set out by the International Liaison Committee on Resuscitation:

Unconcious adult patients with spontaneous circulation after out-of-hospital cardiac arrest should be cooled to 32 °C to 34 °C for 12 to 24 hours when the initial rhythm is ventricular fibrillation (VF).

It has also been suggested, currently without evidence, that hypothermia may be beneficial for other rhythms or for in-hospital cardiac arrest. It is not recommended in patients with severe cardiogenic shock, life-threatening arrhythmias, pregnant patients or those with a coagulopathy.³¹

The uptake of these recommendations is variable. Recently, in the UK, a survey suggested that only 26% of units currently undertook therapeutic hypothermia after cardiac arrest. The main reasons for not using therapeutic hypothermia appeared to be due to logistical or resource issues and a perceived lack of evidence or lack of consensus within individual intensive care unit teams.³²

There is currently insufficient evidence to recommend therapeutic hypothermia in children resuscitated from cardiac arrest. Neonatal animal studies, however, have shown promising results with regard to neuroprotection offered by posthypoxic hypothermia, although it has been shown that the protective effects of hypothermia are lost without adequate sedation.³³

ANAESTHETIC AGENTS

Barbiturates have shown convincing benefit, especially for focal ischaemia in many animal species. In one small clinical study of induced barbiturate coma during coronary bypass surgery, there was a reduction in focal deficits.³⁴ Barbiturate-mediated neuroprotection was initially attributed to suppression of cerebral metabolic rate but more recently to redistribution of CBF to injured areas, the blockade of glutamate receptors and sodium channels, inhibition of free radical formation and potentiation of GABAergic activity.³⁵ At present barbiturates are not commonly used either in coronary bypass surgery or in situations of global ischaemia. Barbiturates are now less commonly used in head-injured patients but can play a role in those patients who have intractable intracranial hypertension.

Propofol has been shown to be neuroprotective in vivo, in focal and global models of cerebral ischaemia. It decreases cerebral metabolic rate and hence CBF. It has been shown to have antioxidant properties, potentiate GABA-A-mediated inhibition of synaptic transmission and inhibit glutamate release. It delays neuronal death by being a free radical scavenger, preventing lipid peroxidation and modulating apoptosis-regulating proteins.^35,³⁶ Its side-effects include hypotension with a reducing CPP and hyperlipidaemia when an infusion of 200 μg/kg per min is used to produce burst suppression.³⁷ This latter problem has been lessened by the introduction of a more concentrated formulation.

Inhalational agents have significant neuroprotective effects but the precise mechanism by which they reduce cerebral injury is unclear.³⁶ It is possible that isoflurane may attenuate excitotoxicity by inhibiting glutamate release and its postsynaptic responses at both anaesthetic and EEG burst supression concentrations. The neuroprotection provided by volatile agents may also be attributable to their effect at GABA-A receptors and an ability to reduce the sympathetic vascular response to ischaemia. Certainly their effect on reducing cerebral metabolic rate is not sufficient to explain their neuroprotective properties.^35,³⁶ These potential benefits in adults are offset by recent studies which have suggested that in neonatal animal studies there may be disturbing increases in cerebral apoptosis, although this has yet to be confirmed. There is an increasing interest in xenon, which presently looks promising.

Nitrous oxide exhibits the neuroprotective and neurotoxic features of an NMDA antagonist. Studies, however, have shown that, when combined with an opioid, e.g. fentanyl, its neuroprotective effect during incomplete cerebral ischaemia is still inferior when compared to volatile agents.³⁶

Midazolam reduces the cerebral metabolic rate for oxygen, CBF and volume. It does not produce burst suppression or an isoelectric EEG, even in large doses.

Neuromuscular blockade is often used in head-injured patients to prevent any coughing on the tracheal tube and subsequent rise in ICP. Their use is not associated with better outcome despite the improvements in ICP control.

CALCIUM ANTAGONISTS

The influx of calcium from the extracellular space and from intracellular organelles has been implicated as the common mediator of cell death from a variety of causes. Calcium antagonists were among the first neuroprotective agents studied to prevent cerebral ischaemia. Despite the effects seen in animal models, human studies in both global and focal ischaemia have been disappointing. Two large trials of intravenous nimodipine in patients with acute ischaemic stroke were terminated early as neurological and functional outcome were significantly poorer in the nimodipine group.^38,³⁹ A close relationship was found between a reduction in diastolic and mean blood pressure in the group treated with nimodipine and an unfavourable neurological outcome. A review of 29 randomised acute stroke trials involving calcium antagonists concluded that the use of calcium antagonists could not be justified in patients with ischaemic stroke and that, although the published trials showed no overall effect on death and dependency, unpublished trials were associated with a statistically significant worse outcome.⁴⁰ Nimodipine, however, has become the standard prophylactic treatment for cerebral vasospasm after subarachnoid haemorrhage, with a consequent decrease in cerebral infarction and better patient outcome.⁴¹ Benefits appear to be due to an effect on smaller penetrating vessels not seen by angiography, or a neuroprotective effect at the cellular level, rather than cerebral vasodilatation identifiable by angiography.

Nimodipine was shown to be neuroprotective in head-injured patients with traumatic subarachnoid haemorrhage.⁴² Regrettably, further studies to test the neuroprotective effect of nimodipine in severely head-injured patients with traumatic subarachnoid haemorrhage (including the multicentre study, HIT IV) failed to confirm the beneficial effects of nimodipine. A small group of patients with a Glasgow Coma Score < 9 did appear to be a have a better outcome in the nimodipine-treated group;⁴³ however, a recent systematic review concluded that there was no beneficial effect on outcome.⁴⁴ Its use therefore remains contentious.