CHAPTER 346 Circulatory Arrest with Deep Hypothermia
Most aneurysms are amenable to endovascular or traditional microsurgical treatment with general neurovascular anesthesia. However, there is a small subset of aneurysms that present the surgeon with an unusually complex technical challenge, and pose a higher risk of surgical morbidity and mortality when traditional techniques are employed.1 Large size, calcified, or atherosclerotic walls, intimacy with critical perforator branches, and the incorporation of afferent or efferent arteries in the dome, are characteristics that often preclude safe clipping while the aneurysm remains tense with high-pressure flow. Circumferential dissection or freeing of adhesions may be impossible with a full, tense dome. Further, calcified or thrombosed aneurysms may be impossible to occlude with a clip without first removing thrombus or calcified components from inside of the aneurysm.
Deep hypothermic circulatory arrest for aneurysm clipping was first employed more than 50 years ago. Initially, the technique carried an unacceptably high incidence of systemic complications, limiting its application to neurosurgery.2 Over time, however, the refinement of bypass techniques via evolving cardiac procedures has allowed for safer employment of circulatory arrest for aneurysm surgery. Coupled with a better understanding of cerebral physiology under hypothermic conditions, these techniques have enabled neurosurgeons to approach otherwise untreatable complex cerebral aneurysms with acceptable morbidity and mortality. Deep hypothermic circulatory arrest is still a relatively high-risk procedure, however, and must performed for appropriate indications at centers able to provide sophisticated neurosurgical and cardiothoracic care. The morbidity and mortality must be weighed against the surgical risks of treating complicated intracerebral aneurysms with traditional techniques.
Temporary Clipping: Limitations of Focal Circulatory Arrest
The utility and complications of temporary clipping for focal circulatory arrest have been explored extensively since its inception in 1961.3–13 Though all cerebral arteries can withstand temporary occlusion for a brief duration, vascular territories vary in the length of occlusion time they may tolerate before ischemia and/or permanent infarction ensue.6 In some cases, the proximal internal carotid artery can be occluded for an hour or more, whereas the perforator-bearing segments of the proximal middle cerebral artery or the basilar apex may tolerate occlusion for only 5 to 15 minutes.8,12 In one recent series, increasing duration of temporary occlusion was shown to correlate significantly with the development of radiographic or clinical infarction.12 Permanent vessel damage and consequent infarction can result from injury to individual perforators captured in the temporary clip because these vessels are not strong enough to withstand the closing force of the clip itself. Additionally, temporarily trapped arterial segments may be subject to intravascular thromboembolism. Finally, a steal phenomenon may occur if a hole is created in the aneurysm while the vessel is temporarily proximally occluded, further exacerbating ischemic intolerance.
Cerebral Ischemia: Pathophysiology and Protection
Cerebral arterial occlusion, either by design or a pathologic event, is known to produce two distinct zones of tissue immediately after occlusion: (1) a central core of densely ischemic tissue, and (2) a penumbra that receives collateral circulation and thus possesses prolonged viability compared to the core. The ischemic core undergoes anaerobic glycolysis, acidosis, adenosine triphosphate (ATP) depletion, and ultimately dysregulated ion homeostasis. The potassium-sodium pump loses function in the absence of ATP, intracellular sodium accumulates, and eventually chloride and toxic calcium accumulate intracellularly causing acidosis, edema, and cell death. If the penumbral zone continues without perfusion, it too is progressively recruited into the ischemic core. Over time the penumbra is exposed to increasing concentrations of toxic neurotransmitters, such as glutamate and nitric oxide released from the actively infarcting ischemic core, which lead to cellular depolarization and spreading neuronal death.14,15
Broadly, the concept of protection from ischemic injury has evolved to include traditional, well-described measures employed during the ischemic insult, and more recently in experimental models, measures implemented before the ischemic insult that provide endogenously mediated protection, known as “ischemic preconditioning.” Preconditioning is a method by which a noxious stimulus near to but below the threshold of damage is applied to cerebral tissue, and shortly thereafter the tissue develops tolerance or resistance. So when the same noxious stimulus is applied at or above the threshold of damage, the tissue is resistant to or protected from injury. Preconditioning has been studied in ischemic stroke, carotid endarterectomy, and myocardial infarction, but has yet to find evidence-supported therapeutic application in these diseases or the ischemia induced in aneurysm surgery.16
Barbiturates are the best studied agent for decreasing brain electrical activity and have been shown to reduce CMRO2 by as much as 50%.17 However, other CMRO2-reducing agents have been used because barbiturates are associated with adverse effects pertinent to hypothermic cardiac arrest. Propofol and isoflurane, for example, are shorter acting and produce less myocardial depression than barbiturates.18 Whichever combination of agents the anesthesiologist uses, one must be mindful that CMRO2 is reduced by only about 50% once electrical activity is eliminated, therefore adding other agents does not effectively lower CMRO2 any further.
Several groups have examined drugs that act by interfering with the toxic events that follow ischemia and deranged ion flux, including free radical scavengers, calcium channel blockers, and glutamate receptor blockers.19–21 Others have examined the beneficial effects of adhesion molecule blockers (anti-ICAM-1 antibody), nonspecific anti-inflammatories (steroids), and rheologic enhancers (mannitol) in preventing progressive microvascular failure and in stabilizing plasma membranes.8 Though these agents tend to be useful for protecting the penumbral zone, they provide little benefit for the already damaged ischemic core.
Hypothermia is one of the few interventions that has proven capable of sufficiently reducing CMRO2 and protecting ionic gradients and structural homeostasis in the setting of prolonged ischemia. Even mild hypothermia down to 33°C has been shown to have demonstrable protective effects, but deep hypothermia to 18°C allows the brain to be totally deprived of blood flow for up to 1 hour without noticeable damage.17
Hypothermia: Cerebroprotective Effects
Hypothermia as a cerebroprotective agent has been used extensively in neurotrauma, cardiac arrest, neurovascular surgery, and cardiac surgery. Deep hypothermia with circulatory arrest has been routinely used in the correction of pediatric cardiac malformations and aortic arch reconstruction.22,23 Ischemic protection provided by hypothermia is attributed primarily to a decrease in metabolic demand (CMRO2), and more recently hypothermia has been shown to reduce proteolysis and excitotoxic damage caused by glutamate toxicity and neuronal calcium influx.17,24,25 Hypothermia has also been suggested to reduce free radical production and eliminate spreading depression in ischemic tissue.26,27 Hypothermia tends to affect the at-risk penumbral tissue but has a minimal effect on the core of infarcted tissue.
Hypothermic tolerance is directly related to the degree of temperature reduction and occurs in the presence of proportional reductions in cerebral blood flow (CBF).28 In cell culture, oxygen consumption has been shown to decrease 50% for every 10°C drop in temperature. However, in vivo reductions in metabolism are even greater (approximately 7% for each 1°C drop in temperature), likely related to hypothermia’s combined effect on basal metabolic rate, electrical activity, and excitatory neurotransmitter and inflammatory cytokine release.29–35 Temporal limitations still exist, however, as humans cannot tolerate deep hypothermia for much longer than 90 minutes. Extended hypothermia results in IQ reduction and neuropsychological deficits as membrane stability becomes threatened.35,36 Therefore, surgeons recommend temporary reperfusion for periods of arrest longer than this.
Hypothermia: Physiologic Considerations
The entire multidisciplinary team, including the neurosurgeon, must understand the widespread systemic effects of profound hypothermia to optimally manage the patient’s intraoperative and postoperative course. Of concern to the anesthesia team is hypothermia’s ability to increase both oxygen and CO2 solubility, which has created controversy among anesthesiologists with regards to temperature-correction of blood gases by adding CO2. Recent evidence suggests it may unnecessarily increase cerebral blood volume, swelling, and intracranial pressure that may be successfully avoided by titrating ventilation to uncorrected gases.37,38 Another major physiologic effect of hypothermia is its role in increasing blood viscosity. Plasma viscosity increases approximately 3% for every 1°C reduction in temperature. The anesthesiologist must therefore counteract the changing viscosity by aggressive isovolemic hemodilution to preserve the integrity of the microcirculatory bed, with the final hematocrit titrated approximately to temperature (i.e., 18°C = HCT of 18).39 Hypothermia may also induce a metabolic acidosis as tissues become relatively underperfused. Metabolic acidosis can then lead to dilation of cerebral vasculature and increased cerebral blood flow, which may be important in the setting of cerebral edema.40,41
Careful glucose monitoring and avoidance of glucose-containing solutions should be maintained during profound hypothermia because hypoinsulinemia and resultant cerebrotoxic hyperglycemia can occur. Endogenous corticosteroid production may be diminished after prolonged hypothermia, necessitating perioperative replacement and continued hypothalamic-adrenal axis surveillance for up to a year. Profound hypothermia can occasionally cause a complement-mediated pneumonitis, which may necessitate prolonged postoperative intubation. Though frank hepatic dysfunction is extraordinarily rare, hypothermia can lead to mild hepatic dysfunction with decreased dilantin metabolism and resultant supratherapeutic levels. Rarely, postoperative renal failure occurs because of transient decreases in glomerular filtration rate, hemolysis, and blood product reactions.41–44 Hemoglobinuria may be managed with osmotic alkalinization of the urine.45,46 Hypothermia may also have profound effects on myocardial contractility and conduction, which persist into the postoperative period. Widening of the PR interval, QRS complex changes, systolic and diastolic dysfunction, sinus bradycardia, and sensitization to pharmacologic challenge have all been reported.47–49
Of greater concern than these rare complications is the more common coagulopathy caused by circulatory arrest techniques and deep hypothermia. The surgeon encounters difficulty with hemostasis because of both platelet dysfunction and slowing of the enzymatic coagulation cascade.50 Fortunately, platelet dysfunction is mostly due to sequestration that resolves within an hour of rewarming, and major clotting factors are mostly unaffected. Heparinization and extracorporeal circuitry account for the remainder of the coagulopathy, and while hypothermia may prolong the anticoagulant effects of the former, its half-life generally is less than 2 hours. As for the latter, removal of the circuits results in an immediate reduction in turbulence and shear-induced red blood cell and platelet destruction. Progressive hemoconcentration during warming also helps to correct the coagulopathy, and warming causes a decrease in plasmin-dependent fibrinolysis.51–53
Although hypothermia is intended to be neuroprotective, neurologic dysfunction can still occur after prolonged, profound hypothermia.54–57 Global hypoperfusion and macroembolization and microembolization may all contribute to the development of neurological injury. Age and preexisting cerebrovascular disease are major risk factors for neurological complications.58 Some recent improvement in neurological complication rates can be found with the use of barbiturate alternatives, which more effectively reduce cerebral metabolism and have fewer cardiovascular effects.38,52,59–61
Deep Hypothermic Circulatory Arrest: Indications
Aneurysm Characteristics
Anatomic Location: Parent Artery
Aneurysms of the posterior circulation have predominated in previously published series of aneurysms treated with hypothermic circulatory arrest.62–64 Some argue that advances in surgical approaches, particularly the orbitozygomatic craniotomy and drilling of the anterior clinoid, provide improved access to the internal carotid artery and give the aneurysm surgeon total proximal control,62 negating the need for circulatory arrest. In fact, before our recently published series of 66 patients outlined below, in which 50% of aneurysms treated with circulatory arrest resided in the anterior circulation, the largest series previously published reported only 4 of 58 aneurysms treated were in the anterior circulation.62 We maintain, however, that particularly risky aneurysms of the anterior circulation may be best treated with hypothermic circulatory arrest.
