Figure 63.1. Pathophysiology cascade of ischemic neuronal death associated to secondary brain damage.
AMPA = α-amino-3-hidroxy-5-methylisoxazole-4-propionic acid type of glutamate receptors; ATP = Adenosin-triphosphate; BAX / Bcl-2 = pro and anti-apoptotic proteins in mytochondrial membrane; Ca++ = calcium; Na+ = sodium; NMDA = N-methyl-D-aspartate receptors; ROS = reactive oxygen species; TNF = Tumor necrosis factor.
Apoptosis-mediated dead cells are phagocytized by cerebral microglia. This ends cellular death expansion and avoids response amplification. However, necrotic cells are phagocytized by macrophages, inducing a propagation reaction in the surrounding tissue. For some authors, changing from a necrotic cellular death to an apoptotic one could be a relevant cerebral-protective mechanism, due to its ability to control neuroinflammation spread. However, this is still a point of debate.
Observing this complex process of ischemic neuronal death helps us to understand some key issues. Firstly, different ischemia-mediated cellular death pathways are intimately related or at least, necrosis and apoptosis are. Secondly, strategies that seemed effective blocking necrotic cellular death may not be relevant in the long term with regards to patient outcome, since apoptotic mechanisms induce late cellular death (this is probably the case of anesthetics-mediated neuroprotection). Finally, it is difficult to imagine a single protective measure capable of slowing down all these events, or altering early features induced by ischemic cascade, owing to the fact that many patients arrive at the emergency area several hours after the onset of the ischemic injury.
In addition to negative effects, ischemia also activates repairing mechanisms. Thus, many processes which result harmful at one point in time can be beneficial at another. For instance, MMPs alter BBB in the early phases of an ischemic event, and therefore inhibition is beneficial at that moment. However they are associated with neurogenesis and angiogenesis later on, and consequently, late treatments based on MMPs-inhibitors might delay reparation mechanisms.
63.3 Cerebral Protection Goals
During early resuscitation, surgery or Intensive Care Unit (ICU) admission, our interventions in an acute brain injury, should focus on preventing and treating cerebral ischemia associated with secondary injury. The final purpose is to minimize cerebral damage and maximize neurological recovery. This objective is based on the fact that primary injuries – with the exception of those caused by some surgical procedures – are seldom predictable (traumatism, stroke, cardiac arrest, etc.) and thus, the initial damage cannot be avoided. Furthermore, secondary injury will decisively contribute to final neurological outcome unless effective neuroprotective measures are implemented. Finally, secondary injury consequences have probably been underestimated in the past. There is evidence supporting the hypothesis that cognitive dysfunction associated with mild cerebral injuries may be due to this secondary injury.
As well as precociousness, other factors seem to be of crucial importance for protective measures to be effective: intensity with which a treatment is applied (drugs given at different doses, levels of hypothermia, etc.), duration of the treatment, severity of primary injury and finally, injury extension – focal or diffuse. Moreover, neurons represent only 5% of cells in cortical grey matter, with glia and astrocytes being much more numerous. Accordingly, another field of interest is neuroprotection “beyond the neuron”, focused on the so called “neurovascular unit” (neurons, muscular brain vessels cells, and support cells – astrocytes, oligodendrocytes and microglia).
Those strategies or treatments which have demonstrated major clinical impact and their applicability will be described hereafter, as well as clinical investigation lines which are currently being developed.
63.4 General Neuroprotection Strategies
Three different groups of neuroprotection strategies can be distinguished:
Neuronal loss reduction. This goal can be achieved by increasing tissular oxygen delivery and decreasing metabolic demands. Traditional neuroprotection was based on decreasing brain metabolism, i.e. administering high doses of barbiturates after an ischemic event. Nevertheless, this pathway is neither the only, nor the most relevant in order to establish neuroprotection. Consequently, other strategies aimed at modifying ischemic neuronal death cascades have been developed:
- Glutamate receptor inhibition (NMDA and AMPA).
- Gamma-amino butyric acid (GABA) system stimulation.
- Anti-inflammatory mechanisms.
- Apoptosis inhibition.
- Mitochondrial dysfunction prevention.
- Cellular hyperpolarization.
Ischemic preconditioning. It is an innate protective mechanism which reduces damage induced by ischemia-reperfusion. Basically, it consists of inducing cellular tolerance by provoking ischemic situations previous to a severe injury that otherwise would be lethal. It has been observed that patients who had suffered from transient ischemic events prior to a serious stroke showed better recovery than those who had not. Two different preconditioning types can be distinguished: early and late preconditioning. The former is developed within minutes to hours after the initial insult, and it is independent of protein synthesis, whilst the latter takes place between 12 and 24 hours after the preconditioning stimulus, lasts for 2 or 3 days, and depends on genetic regulation and protein synthesis.
Neuroprotection can be elicited by certain drugs such as inhaled anesthetics, opioids, adenosine, statins, etc.). They interfere with energetic cellular metabolism and are able to reproduce cellular responses similar to those occurring during preconditioning, such as vasodilatation, angiogenesis, neurogenesis reduction in energy consumption, antagonism of NMDA and AMPA glutamate type receptor [7].
Recently, two additional preconditioning pathways have sparked great interest. On the one hand, remote preconditioning can be elicited by applying brief periods of ischemia to tissues with ischemic tolerance, thus protecting vital organs more susceptible to ischemic damage. For instance, skeletal muscle ischemia is a potent preconditioning stimulus for heart and probably, for brain too [8]. It can be performed quickly and inexpensively, requiring only a blood pressure cuff applied in an arm or a lower limb. Some clinical trials are being carried out in order to test the neuroprotective effect of remote ischemic preconditioning on cardiac surgery (NCT 01231789) and for SAH (NCT01110239).
On the other hand, postconditioning entails brief ischemia periods (10 seconds approximately) applied during reperfusion and separated by 30-second intervals. It appears to be a promising strategy after myocardial ischemia, and its role in cerebral protection is being studied, since cellular injury mechanisms are similar. However, results in neuroprotection are less satisfactory than in the heart, so postconditioning might only be applicable to mild or moderate injuries. This fact might due to a greater cerebral sensitivity towards ischemia [9].
Neuronal reparation. It is mainly caused by a genetic upregulation, between 4 and 24 hours after reperfusion, which induces greater production of endothelial growth factor and erythropoietin (EPO). EPO stimulates neurogenesis in animal ischemia-reperfusion models, decreases neuronal damage and improves long-term cognitive outcomes. EPO and EPO-receptors have been detected throughout the central nervous system, as well as an increase in their levels following hypoxic events. EPO crosses BBB (despite a large molecular weight) and has a very interesting therapeutic window in ischemic events. Its effects involve several cellular levels in the brain: it increases tolerance to ischemia and regeneration of astrocytes, improves differentiation, regeneration and survival of neurons; increases regeneration of oligodendrocytes and preserves vascular regeneration as well as the integrity of the BBB following an ischemic vascular event [10].
GABA-A agonists, such as pregnanolone – a steroid with anaesthetic effects – and sevoflurane are other drugs with neuronal regeneration properties.
63.5 Concrete Strategies for Cerebral Protection
Global strategies for neuroprotection have been described in the section above. Nevertheless, from a practical point of view, it is more useful for physicians to know the potential and applications of measures at their disposal.
63.5.1 Control of Physiological Parameters
At present, this is probably the best modality of cerebral protection [11]. Control and maintenance of arterial pressure, oxygenation, hydroelectrolitic balance and CO2 exchange are key points for final outcomes in several neurocritical pathologies; regardless of whether they are focal or diffuse. However, there is less clear evidence that ideal values of these parameters should be generalized for all patients. That is the reason why cerebral monitoring represents a helpful complement. Neuromonitoring – cerebral perfusion pressure (CPP), jugular bulb oxygen saturation (SjO2), tissue oxygen pressure (ptiO2), transcranial Doppler, cerebral microdialysis, electroencephalography, etc. – should be used, where possible, according to each patient’s pathology, in order to optimize physiological parameters. Since the development of specific neurocritical ICUs has been shown to improve the prognosis of many of these pathologies when compared with general ICUs, admitting neurocritical patients into specialized units could be considered a neuroprotective measure.
Hypoxemia and Hypotension
Hypoxemia and hypotension during the acute phase of cerebral injury are associated with poor prognosis and should be treated aggressively [12]. Oxygen therapy should be focused on maintaining normal oxygenation ranges (SpO2>94%) as hyperoxia during reperfusion may stimulate ROS synthesis and oxidative stress. Hyperoxia has also been shown to reduce lactate levels in cerebral microdialysis, but not the lactate/piruvate rate, which is a more reliable indicator of the energetic state. In addition, lactate may be used as an energetic substrate in case of severe metabolic depletion in the brain. These findings explain why hyperoxia may be harmful in this setting.
Hypotension induces a global reduction in CPP and cerebral blood flow. The initial objective should be to maintain CPP within a target range of 60 to 70 mmHg. Excessive fluid administration may be harmful, due to increased cerebral swelling in case of BBB disruption. Furthermore, arterial hypertension may contribute to vasogenic cerebral swelling in the case of cerebral autoregulation abnormalities, which are frequently established following severe acute injuries. Nevertheless, mild hypertension may be recommended after SAH secondary to cerebral aneurysm rupture, once endovascular coiling or surgical clipping has been performed, in an attempt to avoid vasospasm [13].
Control of arterial blood pressure may reduce cerebral swelling and the risk of hemorrhagic transformation following stroke. Nevertheless, this effect must be balanced against the risk of decreasing perfusion in the penumbra areas. Antihypertensive treatment is discouraged for patients who are non-candidate for thrombolytic therapy, unless diastolic blood pressure is above 120 mmHg or systolic above 220 mmHg. Anti-hypertensive drugs are recommended for patients following thrombolytic treatment, within the first 24 hours, if diastolic or systolic blood pressure are above 110 or 185 mmHg respectively [14]. Despite the lack of clear evidence in spontaneous cerebral hemorrhage, recent studies suggest that maintaining systolic blood pressure below 140 mmHg may reduce hematoma growth [15].
Fluid Management and Transfusion
As a rule, normovolemia is the therapeutic target and, in any case, hypovolemia and hypo-osmolarity should be avoided. Since the brain is very sensitive to secondary injury induced by ischemia, anemia should also be avoided. Hemoglobin levels of 7-8 g/dl appear to be safe for stable patients. On the other hand, acute patients may have a reduced tolerance to anemia. In those cases, the risk-benefit ratio of blood transfusion versus anemia should be assessed [16].
Recent studies have suggested that anemia can foster protective responses in the brain, mediated by ischemic preconditioning, EPO release, β2-adrenergic stimulation, and nitric oxide synthase; but these hypotheses still need clinical confirmation [17]. Monitoring can help in taking individualized decisions for these patients regarding transfusion therapy. Some coexisting diseases (such as coronary disease) may also affect the transfusion threshold.
Concerning osmotic treatment, mannitol is still the most used drug for neurocritical patients, particularly for treating intracranial hypertension [18]. Hypertonic saline solutions are a point of debate: they improve hemodynamic and blood viscosity, reduce endothelial swelling and capillary resistance, but they may cause renal failure, rebound intracranial hypertension, coagulopathy, hypervolemia and electrolytic disturbances [19].
Ventilation
Neurological patients frequently need mechanical ventilation, sometimes for long periods. Ventilation strategies should be aimed at maintaining an adequate gas exchange, minimizing the risk of ventilation-associated lung injury. Cerebral vessels are sensitive to extracellular pH changes caused by ventilation, thus in the range of a paCO2 of 20 to 60 mmHg, cerebral blood flow varies about 3% per mmHg of paCO2. Hypercapnia causes cerebral vasodilatation and, consequently, an increase in cerebral blood volume and blood flow, as well as in intracranial pressure. Hypocapnia (hyperventilation) induces the opposite effects; however, the possibility of critical reductions of blood flow in areas at risk of ischemia should be taken into account. Hyperventilation is effective during brief periods, because extracellular pH tends to normalize a few hours after its instauration, and a rebound phenomenon can even develop once intentional hypocapnia has ceased.
After acute neurological injuries, normocapnia is generally recommended, with short periods of hypocapnia in case of severe neurological compromise, whenever unresponsive to other therapies, and preferentially guided by cerebral oxygenation monitoring (SjO2, ptiO2) and/or microdialysis. Reducing ventilatory dead space, avoiding patient fight against the ventilator, checking permeability of orotracheal tube and procedures directed at increasing lung compliance (i.e. evacuating pleural effusions or severe ascites) are relevant issues for maintaining normocapnia. Hypercapnia is counterproductive and, basically, it should be avoided [20], although some recent investigations have suggested some neuroprotective effects of mild hypercapnia after global cerebral ischemia, possibly via the activation of the hypothalamic-pituitary-adrenal axis, exertion of anti-inflammatory and antioxidant effects, and release of neurotransmitters. However, the real role of hypercapnia in brain injuries remains to be clarified, since deleterious effects may outweigh the possible benefits [21].
There is controversy regarding the use of protective ventilation in patients with neurological injuries, based on a low fraction of inspired oxygen, low tidal volumes and high positive end-expiratory pressure (PEEP). This pattern is often associated to hypercapnia and decreased venous return – the latter mediated by high intrathoracic pressures –, both of which may increase intracranial pressure. PEEP can provide cerebral benefits when its value is lower than intracranial pressure and it enhances pulmonary compliance, thus improving CO2 exchange -via alveoli recruitment- and contributing to cerebral blood flow redistribution (mainly venous). Nevertheless, if PEEP provokes pulmonary overdistension, the net effect can be deleterious [12]. In case of “conflict of interest” between the brain and the lung, putting cerebral protection before protective ventilation is highly recommended [22].
Glucose
Hyperglycemia could simply be a marker of severity of the brain injury; however, without an adequate oxygen supply, its presence increases anaerobic metabolism thus generating acidosis and worsening ischemic injury. Some clinical trials carried out in critical care populations showed a lower intracranial pressure, fewer complications and even lower mortality, when comparing a tight control of glucose blood level versus a liberal one. Whether glucose value per se is more relevant than the insulin administered to maintain that glucose blood level is still controversial, and so is the ideal glucose value (microdialysis patterns characteristic of metabolic neuronal suffering have been observed in patients with glucose levels considered normal in other critical patients) [23]. The glucose threshold in neurocritical patients might be about 140 mg/dl but no definitive data are available on this point.
63.5.2 Therapeutic Hypothermia
Despite not being a drug, it is considered a medical treatment. Its protective effect is beyond decreasing metabolic requirements since it reduces extracellular acidosis, swelling and the release of excitatory aminoacids and free radicals. Besides, it may play a role in repairing BBB integrity, reducing apoptosis and eliminating intracranial gradients of temperature [24]. Cerebral temperature is usually 1°C above core temperature, but differences up to 2°C among different areas of the brain can be observed [25].
Clinical benefits derived from hypothermia probably depend not only on the type of neurological pathology, but also on the precociousness and speed of cooling, the temperature achieved (the protective effect seems to disappear above 35°C, and the complications rate increases below 32°C), the method used for cooling, the duration of hypothermia – which should be guided by clinical targets such as the control of intracranial pressure – and finally, the rewarming rate (much of the benefit can be lost if the patient is rewarmed too fast, independently of the nature of injury) [26]. Therefore, the experience of the centre where therapeutic hypothermia is instituted may be a critical factor.
Mild to moderate hypothermia (33-34°C) in comatose survivors of out-of-hospital cardiac arrest, induced within hours of resuscitation and maintained for 12-24 hours, has demonstrated a reduction in mortality [27]. One meta-analysis showed that the number of patients needed to be treated (NNT) in order to prevent one death was 6, and NNT in order to obtain an optimized neurological recovery was 7. According to this evidence, the 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care included the following recommendations [28]:
- Comatose (i.e., lack of meaningful response to verbal commands) adults patient, with return of spontaneous circulation (ROSC) after out-of-hospital ventricular fibrillation cardiac arrest should be cooled to 32°C-34°C for 12-24 hours.
- Induced hypothermia may also be considered for comatose adult patients with ROSC after in-hospital cardiac arrest of any initial rhythm or after out-of-hospital cardiac arrest with an initial rhythm of pulseless electric activity or asystole.
- Active rewarming should be avoided in comatose patients who spontaneously develop a mild degree of hypothermia (>32°C) after resuscitation from cardiac arrest during the first 48 hours after ROSC.
- Patient core temperature should be closely monitored after ROSC and interventions should be performed to avoid hyperthermia.
- Additionally, the 2010 European resuscitation council guidelines for resuscitation stated that a child with ROSC but remaining comatose after cardiopulmonary arrest might benefit from being cooled to a core temperature of 32-34°C for at least 24 h. The successfully resuscitated child with hypothermia and ROSC should not be actively rewarmed unless the core temperature is below 32°C. Following a period of mild hypothermia, the child should be rewarmed slowly (at 0.25-0.5°C/h) [29].
Moderate whole-body hypothermia (rectal temperature of 33.5-34.5°C) for 72 hours in term neonates with clinical and electrophysiological evidence of perinatal hypoxic-ischemic encephalopathy attributable to perinatal asphyxia showed a strong neuroprotective effect [30], and substantially – although incompletely – reduces neurological disability [31]. Hypothermia must be started within 6 hours of birth and rewarming should last at least 4 hours
The NABISH study, one of the biggest trials run on moderate hypothermia (32.5-34°C), failed to demonstrate either reduction of mortality or functional improvement in adult patients suffering from severe TBI, except for the subgroup of patients younger than 45 who were hypothermic on admission [32]. The recently published NABISH II clinical trial, which evaluated moderate early hypothermia (33°C) for 48 hours in adults younger than 45, has been cancelled without completion of the estimated sample size due to futility, although the subgroup of patients with drainable mass lesions might benefit for hypothermia effect [33].
The methodology of the NABISH trials has been criticized, especially by those who claim that hypothermia contributes to better outcomes provided that it is used with a concrete therapeutic goal (i.e., control of intracranial pressure), without a pre-established duration and checking adverse effects. A systematic literature review concluded that there may be a trend to therapeutic benefits when using hypothermia in patients with severe head injury, but the recommendations are not strong enough due to a lack of studies with adequate methodological quality. The Cochrane Collaboration corroborated these findings [34] and showed an increased rate of pneumonia in TBI patients treated with hypothermia, especially when it was associated to barbiturates. Results for pediatric severe trauma are also contradictory: it appears to improve intermediate outcomes, such as intracranial pressure or cerebral edema, but it has not shown any definitive benefit in terms of survival. Regarding spinal cord injury, there is not clear evidence to recommend for or against the practice of therapeutic hypothermia as a treatment [35]
Indeed, one of the major limitations of hypothermia as a therapeutic measure is the development of complications, especially infectious ones, but also thrombocytopenia, acidosis, insulin resistance, and pancreatitis, which appear to increase the deeper and the longer the hypothermia is. Consequently, physicians should be aware of these complications and treat them early. At present, at least two large randomized clinical trials for evaluating hypothermia in severe TBI patients are being developed: one in Japan and the European Eurotherm trial. Their results will clarify the role of hypothermia as a cerebral protective measure in this disease.
Concerning stroke, current evidence for hypothermia is still scarce and limited to case series and pilot studies. In these patients, hypothermia presents another limitation to its use, as most of them are conscious at the time of hospital admission and do not tolerate cooling. Furthermore, whether or not the association of hypothermia with fibrinolytic therapy enhances neuroprotection is currently being studied [24].
As far as HSA treatment and surgical aneurysm clipping are concerned, a clinical trial which included 1001 patients (IHAST trial), failed to demonstrate any benefit of intraoperative hypothermia in terms of mortality or neurological recovery. Moreover, an increase in the rate of bacteriemia, close to statistical significance, was found in the hypothermia-treated group, discouraging routine use for this indication [36].
Leaving aside the possible beneficial effect of hypothermia, it seems that hyperthermia in patients with neurological injuries is detrimental. Fever increases metabolic rate, triggers the release of excitatory aminoacids and free radicals, and aggravates the damage of the BBB and proteolysis of cellular cytoskeleton. Besides, fever duration influences both vital and neurological prognosis. Therefore fever should be prevented and treated [37], although each type of injury may have an optimal time-window for fever control.
Because we need a balance between limiting secondary injury and impairing the ability to fight against infections, it seems reasonable to control fever aggressively during the first hours to days after spinal cord injury, intracerebral hemorrhage and TBI. In SAH patients the risk period for hyperthermia consequences is more prolonged, since fever can favour vasospasm even some weeks after the hemorrhage [25]. Pharmacologic interventions (mainly acetaminophen), external cooling, maintaining infections surveillance and avoiding shivering can be an appropriate approach to fever prevention and management.
63.5.3 Thrombolysis, Anticoagulants and Antiplatelet Drugs
The main priority in the treatment of stroke is restoring normal cerebral blood flow. The only widely accepted treatment is reperfusion, mediated by intravenous administration of recombinant tissue plasminogen activator (r-tPA) at a dose of 0.9 mg/kg, with a maximum of 90 mg (10% administered as a bolus and the remainder as a continuous infusion for 1 hour). Nevertheless, few patients admitted to hospital with stroke may benefit from reperfusion due to the strict inclusion criteria for therapy that excludes the majority of patients from effective therapy [38] (Table 63.1).
