Tstcsad

Etpapteh1(iIccamare

cf

s1

pfpbr

6

REVIEW ARTICLESWilliam C. Oliver, Jr, MD

Paul G. Barash, MDSection Editors

Perioperative Management of Deep Hypothermic Circulatory Arrest

Marina Svyatets, MD, Kishore Tolani, MD, Ming Zhang, MD, PhD, Gene Tulman, MD, and

Jean Charchaflieh, MD

tairwe

bisacmrpf

mdtretnaahcnilrt1tdut2gglt

HE GOAL OF this article is to provide a review of deephypothermic circulatory arrest (DHCA), covering major is-

ues including pathophysiology of ischemic injury, organ protec-ion (both pharmacologic and nonpharmacologic), temperatureontrol, and perfusion methods. The purpose of this review is toerve as a resource for board review or as introductory material forcardiac anesthesia rotation. The reference list should help in

irecting the reader for more in-depth review of particular areas.The use of therapeutic hypothermia dates back to the ancient

gyptians, Greeks, and Romans.1 In modern times, the use ofherapeutic hypothermia progressed from observation case re-orts to animal studies to clinical use in children and thendults. In 1945, Fay2 provided observational reports of thera-eutic use of hypothermia in patients with severe cerebralrauma. Subsequently, in 1950, Bigelow et al3 reported anxperimental study in dogs that suggested a therapeutic role forypothermia for cerebral protection during cardiac surgery. In959, Drew et al4 reported the use of profound hypothermia12° C, nasopharyngeal) with circulatory arrest (up to 1 hour)n children undergoing surgical repair of the tetralogy of Fallot.n 1975, Griepp et al5 described the use of deep hypothermicardiopulmonary arrest (DHCA, 14-18°C) as a method forerebral protection during the prosthetic replacement of theortic arch.5 Later, the use of DHCA was extended into otherajor vascular surgeries such as the repair of thoracoabdominal

ortic lesions, clipping of giant and complex cerebral aneu-ysms, and resection of renal carcinoma with tumor thrombusxtending into the inferior vena cava or atrium.

DHCA provides 2 clinical benefits. The circulatory arrestomponent provides a bloodless surgical field without the needor the use of intrusive clamps and cannulae. The deep hypo-

From the SUNY Downstate Medical Center, New York, NY.Address reprint requests to Marina Svyatets, MD, SUNY Down-

tate Medical Center, 240 East 47th Street, Apt 20E, New York, NY0017. E-mail: Marina.Svyatets@downstate.eduPublished by Elsevier Inc.1053-0770/2404-0020$36.00/0doi:10.1053/j.jvca.2010.02.010Key words: hypothermia, ischemic injury, end-organ protection,

harmacologic blockade of neurotransmitters, antegrade cerebral per-usion, retrograde cerebral perfusion, intraoperative monitoring, tem-erature management, near-infrared spectroscopy, jugular venousulb oxygen saturation, pH-stat, alpha-stat, cooling, rewarming, du-

aation of circulatory arrest

44 Journal of Cardiothorac

hermic component significantly decreases brain metabolismnd oxygen requirements and thus permits a longer period ofnterrupted blood perfusion to the brain. The cerebral metabolicate is related exponentially to brain (core body) temperature,ith the cerebral metabolic rate decreasing by about 50% for

ach 6°C drop in brain temperature.6

Disadvantages of DHCA include increased cardiopulmonaryypass (CPB) time, edema formation, coagulopathy, and alterationn many organ functions including the kidney, the brain, vascularmooth muscles, intestinal mucosa, alveolar epithelium, the liver,nd the pancreas.7,8 Based on reports from 8 major cardiac surgeryenters in the United States, Europe, and Japan, the risk of per-anent neurologic injury after aortic arch surgery using DHCA

anged from 3% to 12%, renal dysfunction from 5% to 14%,ulmonary insufficiency from 5% to 39%, and left ventricularailure or low-cardiac-output syndrome from 7% to 34%.9

Alternatives to the use of DHCA during aortic arch replace-ent are the use of normothermic CPB10 or mild-to-moderate

egrees of hypothermia. These alternatives obviously requirehe use of a perfusion system for the brain, separate from theest of the body, which might increase the risk of cerebralmbolization.11 Advantages of normothermic CPB include lessime restriction for the completion of surgical repair, mainte-ance of cerebral blood flow (CBF) autoregulation, and avoid-nce of many other disadvantages of DHCA.10 Safety reports ofortic arch repair under less-than-deep degrees of hypothermiaave been provided from small case series and nonrandomizedomparative studies. A case series of 6 patients who underwentormothermic (36°-37°C) aortic arch replacement reported nontraoperative or postoperative mortality or neurologic deficit.10 Aarger case series of 26 patients who underwent thoracic aorticepair under mild (34°C) hypothermic circulatory arrest with an-egrade selective cerebral perfusion (ASCP) at 30°C reported onlypostoperative death caused by the rupture of residual descending

horacic aneurysm and 2 cases (7.7%) of permanent neurologiceficit.12 A larger retrospective comparison of 68 patients whonderwent aortic arch repair under 1 of 3 techniques, mild hypo-hermia (28°-32°C) with ASCP, moderate hypothermia (24°-8°C) with ASCP, or deep hypothermia (18°-24°C) with retro-rade cerebral perfusion (RCP), reported no differences among theroups in either hospital mortality (10.3%) or permanent neuro-ogic dysfunction (8.8%). The mild hypothermic ASCP group hadhe advantages of decreased transfusion volume, intubation time,

nd intensive care unit stay.13

ic and Vascular Anesthesia, Vol 24, No 4 (August), 2010: pp 644-655

P

astailaaaatwblcC

latcpciunrclwsgsm

acaaagh

cpctHffllcRt

s

brbmpaidplp

oaiod

pbpoirdmTctSd

simtl(cfo

ircca(aislemslpu

645DEEP HYPOTHERMIC CIRCULATORY ARREST

athophysiology of Ischemic Brain Injury

Circulatory arrest leads to tissue hypoxia, which affects allerobic functions, particularly the production of the energyource, adenosine triphosphate (ATP) molecules. ATP deple-ion leads to the failure of energy-dependent cell functions suchs the Na�-K�-ATPase pump. This failure is most detrimentaln neural tissues because electrolyte disruption leads to depo-arization dysfunction and, ultimately, cellular structural dam-ge. Failure of the Na�-K�-ATPase pump leads to intracellularccumulation of Na� and Cl�, which leads to cellular swellingnd excessive neuronal depolarization. This depolarization causesn influx of Ca� ions, which activates phospholipases, resulting inhe production of free fatty acids, particularly arachidonic acid,hich leads to hydrolysis of mitochondrial and plasma mem-ranes. During reperfusion, arachidonic acid is further metabo-ized to prostaglandins, thromboxane, leukotrienes, and free radi-als. All these reactions result in an additional accumulation ofa� ions in the cytoplasm.Excessive neuronal depolarization leads to the excessive re-

ease of neuronal excitatory amino acids such as glutamate andspartate. These amino acids are present in excitatory presynapticerminals throughout the brain and are essential for memory,ognition, and consciousness. Glutamate and aspartate are therimary messengers used by neurons for interneuronal communi-ation. After release into the intercellular space, glutamate rapidlys converted to glutamine and then re-enters the neuron ready to besed for the next message. Under normal conditions, powerfuleuronal and glial uptake systems rapidly remove synapticallyeleased excitatory amino acids from the extracellular space. Anyause that interrupts conversion of glutamate to glutamine willead to the accumulation of glutamate in the intercellular space,hereas in increasing concentration it acts as a potent neurotoxic

ubstance. During ischemia, there is insufficient ATP available forlutamine-glutamate conversion and neuron re-entering. Exces-ive neurotransmitter accumulation in the interneuronal spacesay lead to neuronal injury and death.During ischemic conditions, glucose is metabolized in an

naerobic way to lactate, which accumulates in the neurons andauses the development of intracellular acidosis, cell swelling,nd denaturation of proteins and enzymes. A decrease in pH islso a potent stimulus for the release of the glutamate andspartate. The process is accelerated in the presence of hyper-lycemia, and there is ample clinical evidence to suggest thatyperglycemia compounds ischemic cerebral injury.All events in the depolarization phase are reversible, and

urrent clinical protective methods are aimed at delaying orreventing the sequence of these events. Hypothermia andontinued antegrade perfusion are the most effective measureso maintain aerobic glycolysis in the presence of reduced flow.ypothermia and retrograde cerebral perfusion (RCP) are ef-

ective in delaying the depletion of ATP in the zero antegradeow state. Circulatory arrest helps to reduce anaerobic glyco-

ysis and accompanying acidosis by eliminating continued glu-ose supply to fuel the pathway. The trickle flow supplied byCP supplies substrate to maintain anaerobic glycolysis, yet at

he same time may help to remove acid metabolites.14

The collapse of the neurotransmitter transport mechanism

tarts the vicious cycle that constitutes the second phase, the p

iochemical cascade. Excessive activation and the presynapticelease of the excitatory amino acids cause neuronal deathy 2 mechanisms: immediate and delayed. In the immediateechanism, glutamate activates postsynaptic N-methyl-D-as-

artate (NMDA) and alpha-amino-3-hydroxy-5-methyl-4-isox-zolepropionate receptors, leading to Na� and Cl� influx andncreasing cellular edema and, ultimately, membrane lysis andeath. In the delayed mechanism, the activated NMDA receptorromotes the influx of Ca, leading to the activation of phospho-ipases and proteases with formation of free radicals, lipideroxidation, and cell death.The inability to restore calcium homeostasis and the turnover

f cytoskeletal proteins lead to progressive cellular dysfunctionnd apoptosis. Apoptosis usually occurs in zones of borderlineschemia and is an energy-requiring process, whereas necrosisccurs in conditions of complete ischemia and is not energyependent (Fig 1).There are some promising experimental pharmacologic ap-

roaches (neurotransmitter-antagonists, neurotransmitter-receptorlockers, and calcium channel blockers) that aim to modify orrevent the events of the biochemical cascade. Glutamate antag-nists have been shown to be protective in animals after ischemicnjury to the brain.8 However, there is no practical pharmacologicemedy currently ready for clinical application for brain protectionuring ischemia. The search for effective inhibitors of neurotrans-itter release and neurotransmitter receptor blockers continues.14

here is experience with calcium channel blockers with mixedlinical results.14 Amino steroids show promise in countering theoxic effects of free fatty acids, especially arachidonic acid.14

uppression of apoptosis offers a new venue for the prevention ofelayed neuronal loss.14

The last phase of ischemic injury occurs during reperfu-ion and is known as ischemia-reperfusion (IR) injury. IRnjury involves the generation of oxygen free radicals, theost important of which are superoxide radicals, which at-

ack membranes, leading to the further disruption of intracel-ular organelles and cell death. A period of overperfusionhyperperfusion) also may occur after ischemia (including thataused by hypothermic cardiac arrest), leading to a hyperper-usion injury, including cerebral edema, which can worsen theutcome of ischemic injury (Fig 2).Recently, the endothelium has been shown to play a key role

n the injury suffered after ischemia and reperfusion. Whenendered hypoxic and then reoxygenated, endothelial cells be-ome activated to express proinflammatory properties that in-lude the induction of leukocyte-adhesion molecules, proco-gulant factors, and vasoconstrictive agents.15 Nitric oxideNO) is the key endothelium-derived relaxing factor that playspivotal role in the maintenance of vascular tone and reactiv-

ty. In addition to being the main determinant of basal vascularmooth muscle tone, NO opposes the action of potent endothe-ium-derived contracting factors such as angiotensin-II andndothelin-I. NO inhibits platelet and leukocyte activation andaintains the vascular smooth muscle in a nonproliferative

tate. In addition to NO, endothelium may produce other re-axing factors, including prostacyclin, endothelium-derived hy-erpolarizing factor, bradykinin, adrenomedullin, and C-natri-retic peptide. Endothelial dysfunction leads to the decreased

roduction of or availability of NO and/or an imbalance in the

rtrOa

rwlki

eclafi

icgtrt

H

aHm(Crwpt

ing is

646 SVYATETS ET AL

elative contribution of endothelium-derived relaxing and con-racting factors (such as endothelin-I, angiotensin, and oxidantadicals), which leads to an increase in vascular resistance.16

n the other hand, overproduction and accumulation of NOfter hypothermic arrest has been shown to be neurotoxic.17

The aggregation of both platelets and neutrophils lead to theelease of inflammatory mediators and the initiation of thehole-body inflammatory response. Proinflammatory endothe-

ial cell changes promote widespread leukosequestration, cyto-ines release, an increasing level of tumor necrosis factor,nterleukins, oxygen-derived free radicals, and adherence mol-

Fig 1. Pathophysiologic events dur

sFig 2. Ischemia-reperfusion injury.

cules (E-selectin, P-selectin, and intracellular adhesion mole-ule) throughout the body. This further slows the circulation,eading to a cascade of worsening ischemia and cell injury. Innimal studies, leukocyte infiltration and cytokine-depletingltration seemed to mitigate reperfusion injury in the brain.18

Innate immunity now is emerging as an important mechanismn the inflammatory cascade of IR injury. Necrotic or apoptoticell death produces neoepitopes that are recognized by immuno-lobulin M as pathologic.19 This recognition leads to activation ofhe complement system, release of inflammatory cytokines, andecruitment of inflammatory cells that amplify the injury beyondhat caused by the intracellular process alone (Fig 3).

METHODS OF END-ORGAN PROTECTION DURING DHCA

ypothermia

Hypothermia acts by reducing the metabolic rate of the brainnd improving the balance between energy supply and demand.ypothermia reduces cerebral blood flow (CBF) in a linearanner, but the decrease in cerebral metabolic rate of oxygen

CMRO2) is not exactly linear. On average, the reduction inMRO2 is about 7%/1°C. Between 37°C and 22°C, CMRO2 is

educed by about 5%/1°C, and then the reduction accelerateshen CMRO2 reaches 20% at 20°C and 17% at 18°C, at whichoint about 60% of patients achieve electrical silence on elec-roencephalography (EEG) (Fig 4).20

However, animal models of global cerebral ischemia have

chemic injury from circulatory arrest.

hown a protective effect (no injury after 20 minutes of ischemia)

dacla

P

fsciepctocb

8f2m

fheapgtg

ecfhbaivDa

niwarcl

sion i

647DEEP HYPOTHERMIC CIRCULATORY ARREST

uring the use of mild (33°C) hypothermia.21 It could be due todditional mechanisms including halting the ischemic injuriousascade, reducing glutamate excitotoxicity, suppressing intracel-ular calcium influx, decreasing formation of oxygen free radicals,nd increasing gamma-aminobutyric acid release.22

harmacologic Protection

Many pharmacologic interventions have been proposedor organ protection during DHCA. Animal studies haveuggested a beneficial effect of barbiturates, steroids, anti-onvulsants, lidocaine, calcium channel blockers (nimodip-ne), and antagonists to the glutamate receptor subtypes. How-ver, there is little conclusive evidence of benefits throughrospective, randomized, controlled clinical trials. As a result,linical practice varies widely in regard to agents, doses, andiming of administration. A postal survey was sent to membersf the Association of Cardiothoracic Anesthetists (UK) to as-ertain current practice in the use of pharmacologic agentsecause cerebral protective agents during DHCA showed that

Hypoxia

ROS

PARS

EndCel

Ions

KinasesProteasesLipases

Cytoskeletal structures

Membrane lipid, protein

Hypoxia

ROS

PARS

EndCel

Ions

KinasesProteasesLipases

Cytoskeletal structures

Membrane lipid, protein

Fig 3. Inflammatory cascade of ischemia-reperfu

mFig 4. The effect of temperature on CMRO2.

3% of respondents used some form of pharmacologic agentor cerebral protection; 59% of respondents used thiopental,9% used propofol, and 48% used a variety of other agents, theost common of which were steroids.23

Barbiturates act by reducing CMRO2, CBF, free fatty acids,ree radicals, cerebral edema, and seizure activity. Barbituratesave been studied extensively in animal models of focal isch-mia with varying degree of success. Nussmeier et al24 weremong the first to report beneficial effects of thiopental in therevention of neuropsychiatric complications after cardiac sur-ery, but a similar study by Zaidan et al25 could not substantiatehese findings. Trials of barbiturates as protective agents inlobal ischemia failed to show an improvement in outcome.26

It has even been suggested that barbiturates may jeopardize thenergy state of the brain in DHCA patients.23 Despite the lack ofonclusive evidence of neuroprotection, barbiturates still are usedor that purpose in clinical practice during DHCA. Barbituratesave been shown to be protective in incomplete focal ischemiaecause of multiple emboli such as those seen during CPB.23 Inddition, they may be helpful in brain protection during rewarm-ng after DHCA, particularly in the early phase, when jugularenous oxygen desaturation indicates decreased oxygen delivery.isadvantages of large doses of barbiturates include delayed

wakening and myocardial depression.27

Steroids, in particular dexamethasone and methylpred-isolone, counteract the systemic inflammatory response dur-ng and after CPB by decreasing proinflammatory cytokines,hich are thought to play a role in brain ischemic injury as well

s myocardial depression and �-adrenergic desensitization. Ste-oids previously have been shown to improve neurologic out-ome in DHCA patients.28 However, high-dose steroids mightead to an increased risk of sepsis and an alteration in glucose

IgM

ousy

Complement

InflammatoryTissue injury

Cytokineschemokines

Inflammatorycells

IgM

ousy

Complement

InflammatoryTissue injury

Cytokineschemokines

Inflammatorycells

njury. (Color version of figure is available online.)

ogenl injurogen

l injur

etabolism. Glucose control (110-180 mg/dL) should be used

fh

laeaids

inataSchidlbimibrliurcwptcs

caurccr

oetn

miCUtmi

csbcvpmdmcet

hgnuaigsbet

A

b

648 SVYATETS ET AL

or patients receiving glucocorticoids to avoid postoperativeyperglycemia.Mannitol is an osmotic diuretic that protects the kidney by

owering renal vascular resistance, preserving tubular integrity,nd reducing endothelial cell edema. It also reduces cerebraldema and scavenges free radicals, thus reducing tissue dam-ge.29 Furosemide blocks renal reabsorbtion of sodium andncreases renal blood flow, probably mediated by prostaglan-ins. The combination of furosemide and mannitol has beenhown to preserve renal function in ischemic conditions.30

Hyperglycemia might worsen neurologic injury by increas-ng tissue lactic acidosis. Insulin has been shown to have aeuroprotective effect against such injury.31 A retrospectivenalysis of patients undergoing aortic arch surgery revealedhat hyperglycemia more than 250 mg/dL was associated withn adverse neurologic outcome.32 The Society of Thoracicurgeons (US) have laid down guidelines for intraoperativeontrol of blood glucose during adult cardiac surgery becauseigher glucose levels during surgery were found to be anndependent predictor of mortality in patients with and withoutiabetes mellitus.33 As per Society of Thoracic Surgeons guide-ines, a glucose level of �180 mg/dL in diabetic patients shoulde treated with single or intermittent doses of intravenousnsulin. However, if the level is persistently more than 180g/dL, a continuous insulin infusion should be started. If an

nsulin infusion is started in the preoperative period, it shoulde continued in the intraoperative and early postoperative pe-iods to keep the level below 180 mg/dL. The blood glucoseevel should be monitored every 30 to 60 minutes duringnsulin infusion with more intense monitoring (every 15 min-tes) during the administration of cardioplegia, cooling, andewarming. In patients with no history of diabetes, blood glu-ose levels higher than 180 mg/dL should be treated similarlyith single or intermittent doses of intravenous insulin. In suchatients, when insulin infusions are used intraoperatively, cau-ion should be exercised regarding the possibility of hypogly-emia in the postoperative period, and an endocrinology con-ult should be obtained.33

Because Ca�� ions play a major role in the ischemicascade, several studies have examined the role of calciumntagonists as neuroprotective agents.34 Nimodipine, which issed for vasospasm prophylaxis after subarachnoid hemor-hage, has been shown to have some efficacy in improvingognitive outcome after CPB but was associated with increasedomplication rates (hypotension) in patients undergoing valveeplacement, requiring termination of the trial.35

Magnesium, another Ca�� channel blocker, showed evidencef protection against hypoxia in the rat hippocampus.36 This can bexplained by magnesium-induced blockade of both voltage-sensi-ive and NMDA-activated neuronal Ca�� channels, whereasimodipine only blocks voltage-sensitive channels.36

Lidocaine selectively blocks Na� channels in neuronalembranes. In a dog model, high doses of lidocaine induced

soelectric EEG, indicating a pronounced reduction inMRO2.37 In this respect, it mimics the effect of hypothermia.nlike barbiturates, lidocaine can reduce CMRO2 by an addi-

ional 15% to 20%. In this dog model, lidocaine at doses of 4g/kg before DHCA and 2 mg/kg at the start of reperfusion

mproved neurologic deficit scores in the treatment group in

omparison to placebo.37 In a human study, a continuous infu-ion of lidocaine (4 mg/min) during and after CPB resulted inetter short-term cognitive outcome.38 However, a more recentlinical trial of lidocaine in cardiac surgery patients foundariable effects depending on study population. In diabeticatients, there was an association between higher total dose (35g/kg) of lidocaine and increased postoperative neurocognitive

ecline. In nondiabetic patients, lidocaine doses of �42.6g/kg were associated with an improvement in cognitive out-

ome 1 year after surgery. Lidocaine did not decrease periop-rative cytokine response. These findings suggest that the pro-ective effects of lidocaine need to be evaluated further.39

Dexmedetomidine, a selective �2-adrenoreceptor agonist,as been shown in rats to be neuroprotective in both focal andlobal ischemia.40 The inhibition of ischemia-induced norepi-ephrine release might be associated with these effects, partic-larly in the hippocampus. Acadesine, an adenosine-regulatinggent, has been shown to mitigate the effect of reperfusionnjury and decrease stroke rates after coronary artery bypassraft surgery.41 Remacemide, a glutamate antagonist, has beenhown to have neuroprotective effects during coronary arteryypass graft surgery.42 However, there is no sufficient currentvidence to support the clinical use of this drug for neuropro-ection during DHCA42 (Table 1).

djunctive Nonpharmacologic Protection

Recent research has focused on adjunctive methods of cere-ral protection, which may augment the safety of DHCA.43 To

Table 1. Proposed Mechanisms of Action of Potentially

Neuroprotective Pharmacologic Agents

Pharmacologic Agent Proposed Mechanism

Barbiturates Reducing CMRO2, CBF, free fatty acids,free radicals, and cerebral edema.Protective in focal ischemia.

Steroids Decreasing proinflammatory responseMannitol Reducing cerebral edema, scavenging

free radicals, protecting the kidneysby lowering renal vascularresistance, preserving tubularintegrity, and reducing endothelialcell edema

Furosemide Blocking renal reabsorption of sodiumand increasing renal blood flow

Insulin Controlling hyperglycemia, preventingintracellular acidosis

Calcium channelblockers

Blockade of voltage-sensitive andNMDA-activated neuronal Ca2�

channels, decreasing calcium influxinto cytoplasm

Lidocaine Selective blockade of Na� channels inneuronal membranes, reducingCMRO2

Dexmedetomidine Inhibition of ischemia-inducednorepinephrine release, protective inboth focal and global ischemia

Remacemide Glutamate antagonistAcadesine Mitigates the effects of reperfusion

injury

�-Blockers Decreasing inflammatory response

dbfhbmcatbca

clammapiC

ddmpDmc

ipopfflaoarfpt

rpasabmSamubta

(tom

nAmcp

cbpiTasna1epsdptjuptTvarttoRo

649DEEP HYPOTHERMIC CIRCULATORY ARREST

iminish cerebral ischemia time, selective perfusion of therain during deep hypothermia has been implemented in theorm of both RCP or antegrade cerebral perfusion (ACP). ACPas been thought to be superior to RCP for cerebral protectionecause it achieves near-physiologic brain perfusion with ho-ogenous distribution of blood and may extend the safe time of

irculatory arrest.43 RCP, on the other hand, offers only a trivialmount (10%-20% of normal perfusion volume) of bloodhrough brain capillaries, which is thought to meet the meta-olic demands during deep hypothermic conditions, enhanceserebral hypothermia, and might decrease neural intracellularcidosis.44-46

ACP could be hemispheric (through the right axillary, sub-lavian, or innominate artery) or bihemispheric (by adding theeft common carotid artery). There is no consensus regardingppropriate levels of pressure and flow. A flow of 10 to 20L/kg/min is used in a majority of institutions and adjusted toaintain the pressure between 40 and 70 mmHg in the right radial

rtery or 60 to 70 mmHg in the carotid arteries.47,48 Higher-ressure ACP, although theoretically appealing, is associated withncreased CBF, elevated intracranial pressure, high post bypassMRO2, and poorer neurobehavioral recovery.49

Advantages of ACP include a better ability to meet theemands of brain metabolism, flushing brain metabolic wasteuring ischemia, and better control of brain temperature. ACPay obviate the need for deep hypothermia, thus reducing

ump time and complications related to deep hypothermia.13

rawbacks of ACP include the risks of arterial wall dissection,alperfusion, embolism of atheromatous plaque or air, the

luttered operative field, and the cumbersome procedure.RCP is achieved by cannulating the superior vena cava and

nstituting flow rates of 300 to 500 mL/min to maintain a meanressure of 25 to 35 mmHg. Occlusion of the inferior vena cavar snaring of the superior vena cava often is done to preventreferential flow of blood to the inferior vena cava. RCP allowsor deep and homogenous cooling of the brain and helps inushing solid particles, air bubbles, and metabolites from therteries, thus reducing embolic phenomena and delaying onsetf acidosis in the ischemic brain. RCP is more efficacious in thebsence of atherosclerotic disease.50 RCP has been reported toeduce mortality rates from 14.8% to 7.9% and stroke ratesrom 6.5% to 2.4% during DHCA.51 Other investigators re-orted similar findings in reducing stroke rates from 9% to 3%hrough the use of RCP during DHCA.52

Cerebral hypothermia during RCP is achieved through arte-iovenous shunts to overcome any deficiencies in transcapillaryerfusion. Studies have found that most RCP blood is shuntedway from capillaries even during normothermia, and thishunting is increased during deep hypothermia as arteriovenousnd venovenous shunts open.53 The partial perfusion providedy RCP is insufficient to sustain cerebral metabolism, whichight be impaired further with RCP-induced cerebral edema.54

ome investigators suggested that a combination of circulatoryrrest under moderate systemic hypothermia and cold RCPight provide adequate cerebral protection for up to 30 min-

tes.55 Others have suggested the following strategies on theasis of the expected operative procedure and circulatory arrestime: (1) for limited arch replacement with short circulatory

rrest time (30-40 minutes), DHCA alone would be sufficient; t

2) for more extensive repairs that require prolonged circula-ory arrest times, DHCA plus ACP is recommended; and (3) forperations with high embolic risk, DHCA plus RCP is recom-ended56,57 (Table 2).

MONITORING DURING DHCA

Monitoring for adult patients undergoing DHCA includes alloninvasive monitoring according to the standards of themerican Society of Anesthesiologists, invasive hemodynamiconitoring including an arterial catheter and pulmonary artery

atheter, transesophageal echocardiography (TEE), and neuro-hysiologic monitoring.TEE is extremely useful in monitoring many aspects of

ardiovascular function including assessing cardiac functionefore and after DHCA, examining the entire aorta, confirmingroper cannula placement, assessing volume status, detectingntracardiac air, and evaluating the adequacy of surgical repair.emperature monitoring is standard during general anesthesiand is essential during hypothermic techniques. Sites of mea-uring core body temperature include the tympanic membrane,asopharynx, esophagus, urinary bladder, rectum, pulmonaryrtery, and jugular venous bulb. It is preferred to use more thansite for core body temperature monitoring to detect differ-

nces in circulatory distribution. Common sites include theulmonary artery and the urinary bladder. The esophageal siteeldom is used in conjunction with transesophageal echocar-iographic monitoring for safety and efficacy concerns. Tem-erature monitoring at the tympanic membrane might providehe closest assessment of brain temperature.58 When used,ugular venous temperature monitoring would be particularlyseful during rewarming to ascertain the lack of cerebral hy-erthermia because jugular venous bulb temperature is consis-ently greater than that of any other sites including the brain.59

he CPB outflow temperature is the best indicator of the jugularenous blood temperature. When DHCA is intended, circulatoryrrest usually is initiated when the core body temperature haseached around 12° to 15°C. During rewarming, the perfusateemperature is kept at a maximum of 10°C above core bodyemperature and never above 36°C. A urinary bladder temperaturef about 34°C is used as an indicator of adequate rewarming.ewarming to greater temperatures is avoided to prevent danger-us rebound hyperthermia after CPB.60

Table 2. Comparative Characteristics of ACP and RCP as

Adjunctive Methods of Cerebral Protection

Characteristics ACP RCP

Simplicity and ease of application �� �

Adequacy to support cerebralmetabolism �� �

Reduced pump time � ��

Limited manipulation of archbranches �� ��

Reduced embolic load �� ��

Risk of arterial wall dissection,malperfusion, embolism �� ��

Interstitial edema, cerebral edema � ��

NOTE. � and � indicate degree of presence or absence, respec-

ively.

sv

scobvca6tscct�wSii

iI(rcosTacfgl

(rmie

crmtaancacTt

avdwivpt

discmdtibseu1aotihiieilmsEec

gitth

650 SVYATETS ET AL

Neurophysiologic monitoring might include EEG, somato-ensory-evoked potentials, oxygen saturation of the jugularenous bulb (SjO2), and near-infrared spectroscopy (NIRS).

NIRS is a noninvasive monitoring technique that mea-ures regional cerebral oxygen saturation (rSO2) and detectshanges in cerebral oxyhemoglobin, deoxyhemoglobin, andxidized cytochrome aa3 concentration in brain tissue. Inrain tissue, the vascular compartment is predominantlyenous (70%-80% v 20%-30% arterial). Oxygen saturation oferebral venous blood is about 60% versus 98% to 100% in therterial blood. Based on these assumptions, the average rSO2 is0% to 70%. During cardiovascular surgery, decreasing rSO2

rends seem to reliably reflect decreasing cerebral oxyhemoglobinaturation. Levels of rSO2 of �55% are indicative of neurologicompromise and have been associated with adverse clinical out-ome including postoperative cognitive dysfunction.61 During aor-ic arch surgery under DHCA and ACP, decreases in rSO2 to

55%, particularly when longer than 5 minutes, were associatedith the occurrence of postoperative neurologic adverse events.62

ome authors have suggested maximizing rSO2 values beforenstituting DHCA; however, the role of such a method in improv-ng outcome has yet to be shown.63

The earliest NIRS devices to receive Food and Drug Admin-stration approval (mid 1990s) and, thus, the most studied areNVOS (Somanetics Corporation, Troy, MI) and NIROHamamatsu Photonics KH, Hamamatsu City, Japan). A recenteview stated that a majority of high-volume aortic arch surgeryenters are using cerebral NIRS on a routine basis.64 Previ-usly, the same author suggested an algorithm of managingigns of decreased cerebral oxygenation as detected by NIRS.he proposed factors to be addressed and their order of man-gement is as follows: the position of aortic and superior venaava cannulae including head and neck position, cerebral per-usion pressure including mean arterial pressure, arterial oxy-en content, partial pressure of carbon dioxide, hemoglobinevel, cardiac output, and CMRO2.65

NIRS has several limitations that need to be consideredTable 3). It provides information on rSO2 in only a limitedegion of the brain (part of the frontal lobes) and cannotonitor the entire brain. Therefore, rSO2 might fail at detecting

schemia at sites different from monitored ones. The use oflectrocautery may interfere with the NIRS monitoring. The

Table 3. Basic Advantages and Disadvantages of NIRS

Advantages Disadvantages

Noninvasive Cannot monitor entire brainContinuous Electrocautery interferenceReal time Can be affected by variation of cerebral

blood flowInexpensive Can not differentiate between causes of

declining rSO2 valuesPortable, readily

availableCan be affected by different forms of

hemoglobin, excessive bilirubinSafe Can produce normal rSO2 over an area

with dead brain tissueEasy to interpret Interpretation may be confounded by

hypothermia, alkalosis, or

thypocapnia

ontribution of external carotid blood flow and variations in theatio of cerebral arterial/venous blood flow (contaminants)ight result in inaccurate rSO2 values. NIRS cannot differen-

iate between the different causes of declining rSO2 (embolusnd malperfusion). It can be affected by excessive bilirubinnd different forms of hemoglobin. NIRS can produce aormal rSO2 reading when positioned over an area in whicherebral perfusion is absent and the brain tissue is presum-bly dead. The interpretation of absolute rSO2 values may beonfounded by hypothermia, alkalosis, or hypocapnia.herefore, it is essential to follow trends in oxygen satura-

ion changes rather than absolute values.66

Oxygen saturation of the jugular venous bulb (SjO2) has beendvocated as a marker of global cerebral oxygenation.67 Decreasedalues of SjO2 indicate a decreased oxygen supply relative toemand. Values of �50% during rewarming have been associatedith postoperative cognitive decline.67 Increased values of SjO2

ndicate decreased oxygen extraction relative to supply. Increasedalues might be caused by a hypothermic decrease in CMRO2,harmacologic suppression of CMRO2, or severe brain injury tohe point of reduced oxygen extraction.67

Electroencephalographic monitoring provides continuousetection of brain electrical activity. EEG can be used beforenitiating DHCA and during DHCA to document electricalilence, which reduces CMRO2 by about 50%. It is a nonspe-ific monitor of global ischemia, which could be caused byalperfusion, hypotension, or CPB. It is more specific for the

etection of epileptiform activity, which might be seen atemperatures of 30°C or caused by pathologic causes includingschemia. EEG is used to detect burst suppression, which cane induced at temperatures of 24°C or by pharmacologic agentsuch as barbiturates or etomidate. Burst suppression providesvidence of suppression of metabolic brain activity. EEG issed to detect electrical silence, which is achieved at about7°C and which provides ultimate suppression of metabolicnd electric brain activity. Circulatory arrest can be institutednce electrical silence has been present for 3 minutes. Limita-ions of electroencephalographic monitoring include its inabil-ty to reflect the activity of deeper brain structures, such asippocampus and basal nuclei, which are very vulnerable toschemia that affects neurocognitive function.68 Although EEGs likely to reflect changes in energy expenditure necessary forlectrical neuronal transmission, it is unlikely to reflect changesn energy consumption necessary for the maintenance of cel-ular integrity. Therefore, EEG is insufficient as an isolatedethod for ascertaining adequacy of cerebral protection. Somato-

ensory-evoked potentials are generally easier to implement thanEG because it is less prone to external electric noise, less influ-nced by anesthetic drugs, and remains detectable as long asortical activity is present.69

ACID-BASE MANAGEMENT DURING HYPOTHERMIA

Hypothermia alters the results of analysis of arterial bloodases by increasing the solubility of CO2 and O2 in plasma. Thencrease in CO2 solubility decreases the concentration ofhe insoluble portion and, thus, the partial pressure. However,he total content of CO2 in the blood remains the same. Duringypothermia, if a blood sample is taken and warmed to 37°C in

he blood gas analyzer, the CO2 initially dissolved will now

cwhtIt

3tn

tmblicsap

badCm

mmpier

aabqatwl

srdwCbraDc

twcbi

oif

C

eoDamyaomc

es(Hflvdp

T

opdkiDom�itcu

R

cibrrmtbd

di

651DEEP HYPOTHERMIC CIRCULATORY ARREST

ontribute to the partial pressure of CO2 (PCO2) and the PCO2

ill be within the normal normothermic range. If, on the otherand, the value is estimated at the patient’s actual temperature,he PCO2 will be reduced despite similar arterial CO2 content.n addition to its effect on gas solubility, hypothermia decreaseshe metabolic rate and CO2 production.

Maintaining the PCO2 within the normal range in rewarmed7°C blood is called “alpha-stat.” If the PCO2 is corrected tohe patient’s actual temperature and that value is kept within theormal range, the management is called “pH-stat.”Alpha-stat management is aimed at maintaining autoregula-

ion of the brain and cellular enzymatic activity. It does so byaintaining normal acid and blood gas values in the rewarmed

lood. At hypothermic temperatures, the blood remains alka-emic, and this, in addition to hypothermia, increases the affin-ty of hemoglobin to oxygen. Alpha-stat management wouldause relative hypocarbia, which would produce cerebral va-oconstriction and reduce CBF. This would make this approachdvantageous in decreasing embolic load and preventing over-erfusion and subsequent cerebral edema.The pH-stat method aims at maintaining normal values in vivo

y adding CO2, and when rewarmed the blood becomes acidemicnd hypercarbic. The resulting hypercarbia causes cerebral vaso-ilation, increased CBF, and a loss of autoregulation. IncreasedBF in conjunction with reduced CMRO2 allow quick and ho-ogenous cooling of the brain and increased oxygen delivery.The pediatric literature generally supports the use of pH-statanagement during DHCA for providing both cerebral andyocardial protection.70 Clinical studies suggest that pH-stat is

articularly beneficial in cyanotic neonates and infants becauset shifts more CPB flow away from the aortopulmonary collat-ral circulation toward the cerebral circulation, improving ce-ebral cooling and oxygen supply.71

Because it maintains a physiologic coupling between CBFnd CMRO2, the alpha-stat strategy appears advantageous indults in whom the risk of under- or overperfusion within therain is substantial. Cerebral edema, which can be a conse-uence of cerebral overperfusion, is less likely to occur withlpha-stat strategy. However, the preservation of cerebral au-oregulation may attenuate the uneven distribution of blood,hich might occur in patients with an underlying vasculopathy

ike atherosclerosis, hypertension, or diabetes.72

Studies favoring pH-stat revealed significantly fewer epi-odes of jugular venous desaturation upon rewarming andeductions in cerebral arteriovenous glucose and oxygen gra-ients in comparison to alpha-stat.73,74 Prior studies have shownorse neurologic outcomes using pH-stat during hypothermicPB.75 Theoretically, pH-stat management might provide theenefit of excessive cerebral blood flow but with an increasedisk of macro- or microcerebral embolization.76 Empirically, inn animal (pig) model of controlled microembolic load duringHCA, pH-stat was associated with improved outcomes when

ompared with alpha-stat.77

In the absence of definitive data favoring one strategy overhe other, it might be prudent to use a combined strategy inhich pH-stat is used while cooling, thus using the benefits of

erebral vasodilatation to enhance the rate and homogeneity ofrain cooling; followed by the alpha-stat strategy from the

nterval immediately before cessation of circulation to the time i

f rewarming, thus minimizing extracellular acidosis and aim-ng for the preservation of CBF-CMRO2 coupling during reper-usion and rewarming.78

TEMPERATURE MANAGEMENT

ooling

The cooling phase should be gradual, thorough, and longnough to achieve homogenous allocation of blood to variousrgans and to prevent a gradual updrift of temperature duringHCA.79 To achieve a final core body temperature (bladder

nd esophageal) of 10° to 13°C, cooling should last at least 30inutes.80 Rapid cooling might create imbalance between ox-

gen delivery and demand, and it might decrease oxygen avail-bility to the tissues by increasing the affinity of hemoglobin toxygen. This increased affinity combined with extreme he-odilution from the priming solution for CPB might lead to

ellular acidosis before DHCA.Moderate hemodilution leads to improved microcirculation, but

xtreme hemodilution might lead to tissue hypoxia. Studies havehown that intracellular acidosis was not present with a hematocritHct) level of 30%, mild with an Hct of 20%, and severe with anct of 10%.81,82 These studies showed that the cerebral capillaryow was maintained with an Hct of 30% despite increased bloodiscosity during deep hypothermia.82,83 Appropriate levels of Hcturing pre- and post-DHCA would range from 22% to 28%, withroportional relationship to the core temperature.

opical (Head) Cooling

A delay in temperature equilibrium may occur because ofcclusive vascular disease that reduces cerebral perfusion. Iceacking of the skull enhances cerebral hypothermia via con-uction across the skull.84 Ice packing the head, in addition toeeping body temperature around 10° to 13°C, also might helpn preventing an undesirable rewarming of the brain duringHCA. An animal (pig) study showed improved behavioralutcome with head ice packing during DHCA.85 Another ani-al study showed improved recovery of metabolic function by50% in piglets who had their heads packed in ice in compar-

son to those who did not.86 Systems of continuous cooling ofhe head during DHCA recently have been developed. Theyonsist of a cooling cap with an incorporated circuit of contin-ously circulated water at a desirable temperature.87

ewarming

Rewarming increases CBF and the risk of embolization,erebral edema, and hyperthermic brain injury. During rewarm-ng, extracranial sites of temperature monitoring underestimaterain temperature by about 5° to 7°C; therefore, caution isequired to avoid hyperthermic arterial inflow, which mayesult in brain hyperthermia.88 During rewarming, it is recom-ended that the perfusate temperature not exceed core body

emperature by more than 10°C; to stop rewarming when coreody temperature is 36°C (esophageal) or 34°C (urinary blad-er); and for perfusate temperature not to exceed 36°C.60

Relative hypothermia (36°C, esophageal; 34°C, urinary blad-er) might be beneficial to prevent cerebral electrical hyperactiv-ty.89 When available, EEG should be monitored during rewarm-

ng to detect electrical hyperactivity, and when detected it should

blbaci

tAocdaec1taiD

bplsmm

aDpf(iamhb

mtpocCbc

Dd1(nipzrndibtDhDtatf

mwfDgarhebn

bl(hctb

ss

652 SVYATETS ET AL

e treated promptly with deepening the level of anesthesia andowering the temperature. Initial reperfusion with relatively coldlood at low pressures allows washout of accumulated metabolitesnd free radicals and provides substrates for high-energy mole-ules. A period of initial hypothermic perfusion has been shown tomprove neurologic outcome.90

DURATION OF DHCA

A number of biochemical and cellular structural changesake place as the duration of circulatory arrest lengthens.fter 15 minutes of ischemia at 18°C, the recovery ofxygen consumption is impaired, and after 20 minutes,erebral lactate is detected in the effluent blood. The safeuration of circulatory arrest at 15°C was predicted to bebout 29 minutes and at 10°C about 40 minutes.91 If isch-mic tolerance is considered 5 minutes at normothermia, thealculated safe period of circulatory arrest at 18°C would be5 minutes91 (Fig 5). Clinical studies have shown a persis-ent loss of cognitive function (lasting more than 6 weeks)nd deterioration in postoperative cognitive scoring/testingn patients who underwent aortic arch surgery by usingHCA for more than 25 minutes at 10°C.92

Intermittent DHCA with low-flow (“trickle”) CPB haseen used to extend the safe duration of DHCA. Intermittenterfusion is associated with higher oxygen saturation andower lactate levels in the sagittal sinus, indicating a betterupply of energy demands. Metabolic hemostasis is betteraintained when intermittent perfusion is used every 20inutes at 18°C.92

An experimental study using a combination of head ice packingnd low-flow (“trickle”) CPB during DHCA showed (1) thatHCA caused an impairment of cerebral metabolism, directlyroportional to its duration; (2) a better recovery of metabolicunction (improving �50%) in the group with head ice packing;3) a “trickle” of a CPB of only 5 to 10 mL/kg/min during DHCAs superior to a DHCA only application at the same temperature;nd (4) despite achieving a core body temperature of 18°C for 30inutes, the brain requirement for O2 may persist according to a

igher level of temperature because the brain temperature tends toe higher than core body temperature.86

iFig 5. The relationship between the degree of hypothermia and

afe duration of circulation arrest (CA).

Pulsatile flow maintains patent microcirculation at lowerean perfusion pressures than nonpulsatile flow, which allows

he use of lower flow rates. Peak pressures developed duringulsatile flow helps to overcome the critical opening pressuref the capillaries. Pulsatile flow is associated with reducederebral vascular resistance and improved recovery ofMRO2.93 It also might be advantageous in improving thealance between myocardial oxygen supply and demand, espe-ially during rewarming.94

NEUROLOGIC INJURY CAUSED BY DHCA

Neurologic injury is the most troublesome adverse effect ofHCA and CPB, presenting either as transient neurologiceficit (5.9%-28.1%) or irreversible neurologic injury (1.8%-3.6%). Early postoperative mortality markedly is increased18.2%) in patients with neurologic injury, and long-term cog-itive disability is common among survivors.95 A focal deficits usually an embolic phenomenon, whereas a prolonged poorerfusion of the brain may produce necrosis in watershedones. Age, atherosclerosis, and manipulation of the aorta areisk factors for both. Global cerebral ischemia leads to diffuseeurologic deficit, which may be benign and reversible or moreebilitating (seizures, Parkinsonism, and coma). Risk factorsnclude increased duration of circulatory arrest and CPB, dia-etes mellitus, and hypertension. Transient neurologic dysfunc-ion appears to be a marker of long-term cerebral injury.96

eficits of memory and fine-motor function may persist afterospital discharge. Reductions in CMRO2 and the duration ofHCA reduce the risk of neurologic injury.32 The length of

ime on CPB might be a better predictor of postoperative deathnd stroke than the duration of DHCA time.97 Microemboliza-ion during prolonged CPB is likely to be a greater risk factoror stroke than cerebrovascular ischemic time.43

In the pediatric population, evidence of overt brain injuryight be found in up to 10% of patients exposed to DHCA,hereas subtle but detectable neuropsychiatric defects might be

ound in up to 50%.98 In infants undergoing heart surgery,HCA, compared with low-flow CPB, was associated with areater degree of central nervous system perturbation (clinicalnd electroencephalographic seizures and a longer time to theecovery of normal brain activities as assessed by EEG) and aigher level of BB isoenzyme of creatine kinase during thearly postoperative period, but at the time of hospital dischargeoth groups had similar overall incidence of abnormalities oneurologic examination and EEG.99

Measures that clearly have been documented to providerain protection include the following: (1) systemic cooling andocalized ice packing, (2) electroencephalographic silence, and3) continuous or intermittent ACP. Measures that possiblyave a role in brain protection include the following: (1)ontinuous RCP and (2) pharmacologic blockade of neuro-ransmitters. Hypothetical measures include pharmacologiclockade of reperfusion injuries.

CONCLUSION

DHCA is an established technique that is used duringurgical repair of the aortic arch and other major vessels

ncluding the thoracoabdominal aorta, cerebral vessels, and

inmmDntuTetD

sraeoccnvce

OI

c

ps1

c

p1

am

Ap

c

T

aS

ap

hC

pt

evI

i1

f

sc

653DEEP HYPOTHERMIC CIRCULATORY ARREST

nferior vena cava. Circulatory arrest provides the conve-ience of a bloodless surgical field, whereas deep hypother-ia provides significant protection to the brain and otherajor organs against circulatory arrest. Alternatives forHCA include lesser degrees of hypothermia, to the point oformothermia, and lesser degrees of circulatory arresthrough the use of various forms of intermittent or contin-ous perfusion techniques of the brain and other organs.here are both theoretic and experimental evidence of ben-fits and risks of DHCA and each of its alternatives. Al-hough there is no conclusive evidence of the superiority of

HCA or any of its alternatives in affecting overall or v

REN

ulatory arrest. Ann Thorac Surg 64:1639-1647, 1997

tcC

mi

dC

ie

c2

ct5

ct

oA

ca

um2

sr

hb

sF

ca

RJ

pecific outcomes in indicated procedures, there is sufficienteason to base the use of DHCA and/or any of its adjunctsnd alternatives on the complexity of the vascular lesion, thexpected duration of surgical repair, and the expertise of theperative team. Regardless of the degree of hypothermia orirculatory arrest that is being used, advances in monitoringerebral and other organ functions and in pharmacologic andonpharmacologic therapeutic interventions continue to pro-ide tools for improving the outcome of care of patients withomplex vascular lesions. Enhanced understanding of isch-mic injury and IR injury provide further targets for inter-

ention to improve outcome.

REFE

1. Polderman KH: Application of therapeutic hypothermia in the ICU:pportunities and pitfalls of a promising treatment modality. Part 1:

ndications and evidence. Intensive Care Med 30:556-575, 20042. Fay T: Observations on generalized refrigeration in cases of severe

erebral trauma. Assoc Res Nerv Ment Dis Proc 24:611-619, 1943.3. Bigelow WG, Lindsay WK, Greenwood WF: Hypothermia; its

ossible role in cardiac surgery: An investigation of factors governingurvival in dogs at low body temperatures. Ann Surg 132:849-866,9504. Drew CE, Keen G, Benazon DB: Profound hypothermia. Lan-

et 1:745-748, 19595. Griepp RB, Stinson EB, Hollingsworth JF, et al: Prosthetic re-

lacement of the aortic arch. J Thorac Cardiovasc Surg 70:1051-1063,9756. Gega A, Rizzo JA, Johnson MH, et al: Straight deep hypothermic

rrest: Experience in 394 patients supports its effectiveness as a soleeans of brain preservation. Ann Thorac Surg 84:759-766, 20077. Pieter J, van der Starre A: Deep Hypothermia and Circulatory

rrest in Thoracic Aortic Diseases. Berlin, Germany, Springer, 2006,p 101-1088. Haverich A, Hagl C: Organ protection during hypothermic cir-

ulatory arrest. J Thorac Cardiovasc Surg 125:460-462, 20039. Moon MR, Miller DC: Aortic arch replacement for dissection. Op

ech Thorac Cardiovasc Surg 4:33-57, 199910. Touati GD, Marticho P, Farag M, et al: Totally normothermic

ortic arch replacement without circulatory arrest. Eur J Cardiothoracurg 32:263-268, 200711. Kunihara T, Grun T, Aicher D, et al: Hypothermic circulatory

rrest is not a risk factor for neurologic morbidity in aortic surgery: Aropensity score analysis. J Thorac Cardiovasc Surg 130:712-718, 2005

12. Toyama M, Matsumura Y, Tamenishi A, et al: Safety of mildypothermic circulatory arrest with selective cerebral perfusion. Asianardiovasc Thorac Ann 17:500-504, 199913. Kaneda T, Saga T, Onoe M, et al: Antegrade selective cerebral

erfusion with mild hypothermic systemic circulatory arrest duringhoracic aortic surgery. Scand Cardiovasc J 39:87-90, 2005

14. Klein JJ, Ergin A: Principles of cerebral protection during op-rations on the thoracic aorta, in Franco KL, Verrier ED (eds): Ad-anced Therapy in Cardiac Surgery. Hamilton, London, BC Deckernc, 2003, pp 291-303

15. Boyle EM, Pohlman T, Cornejo C, et al: Endothelial cell injuryn cardiovascular surgery: Ischemia-reperfusion. Ann Thorac Surg 62:858-1875, 199616. Verma S, Anderson TJ: Fundamentals of endothelial function

or the clinical cardiologist. Circulation 105:546-549, 200217. Tseng RR, Brock MV, Lange MS, et al: Neuronal nitric oxide

ynthase inhibition reduces neuronal apoptosis after hypothermic cir-

CES

18. Rimpilainen J, Pokela M, Kiviluoma K, et al: Leukocyte filtra-ion improves brain protection after a prolonged period of hypothermicirculatory arrest: A study in a chronic porcine model. J Thoracardiovasc Surg 120:1131-1141, 200019. Zhang M, Carroll MC: Natural Ig M-mediated innate autoim-unity: A new target for early intervention of ischemia-reperfusion

njury. Expert Opin Biol Ther 7:1575-1582, 200720. Kimura T, Muraoka R, Chiba Y, et al: Effect of intermittent

eep hypothermic circulatory arrest on brain metabolism. J Thoracardiovasc Surg 108:658-663, 199421. Busto R, Dietrich WD, Globus MY, et al: Small differences in

ntraischemic brain temperature critically determine the extent of isch-mic neuronal injury. J Cereb Blood Flow Metab 7:729-738, 1987

22. Dorotta I, Kimball-Jones P, Applegate R: Deep hypothermia andirculatory arrest in adults. Semin Cardiothorac Vasc Anesth 11:66-76,007

23. Dewhurst AT, Moore SJ, Liban JB: Pharmacological agents aserebral protectants during deep hypothermic circulatory arrest in adulthoracic aortic surgery. A survey of current practice. Anaesthesia7:1016-1021, 200224. Nussmeier NA, Arlund C, Slogoff S: Neuropsychiatric compli-

ations after cardiopulmonary bypass: cerebral protection by a barbi-urate. Anesthesiology 64:165-170, 1986

25. Zaidan JR, Klochany A, Martin WM, et al: Effect of thiopentaln neurologic outcome following coronary artery bypass grafting.nesthesiology 74:406-411, 199126. Brain Resuscitation Clinical Trial 1 Study Group: A randomized

linical study of thiopental loading in comatose survivors of cardiacrrest. N Engl J Med 314:397-403, 1986

27. Hirotani T, Kameda T, Kumamoto T, et al: Aortic arch repairsing hypothermic circulatory arrest technique associated with phar-acological brain protection. Eur J Cardiothorac Surg 18:545-549,

00028. Shum-Tim D, Tchervenkov CI, Jamal AM, et al: Systemic

teroid pretreatment improves cerebral protection after circulatory ar-est. Ann Thorac Surg 72:1465-1471, 2001

29. Yang MW, Lin CY, Hung HL, et al: [Mannitol reduces plasmaydrogen peroxide free radical in patients undergoing coronary arteryypass graft surgery]. Ma Zui Xue Za Zhi 30:65-70, 199230. Hanley MJ, Davidson K: Prior mannitol and furosemide infu-

ion in a model of ischemic acute renal failure. Am J Physiol 241:556-F564, 198131. Strong AJ, Fairfield JE, Monteiro E, et al: Insulin protects

ognitive function in experimental stroke. J Neurol Neurosurg Psychi-try 53:847-853, 1990

32. Ceriana P, Barzaghi N, Locatelli A, et al: Aortic arch surgery:etrospective analysis of outcome and neuroprotective strategies.

Cardiovasc Surg (Torino) 39:337-342, 1998

Srde

ae

nm

np

dA

pg

dc

bc

agG

to2

cA

uA

r

oT

p6

ts

cJ

ot2

p1

cT

eJ

nd

up

t

raa

tt

JS

LO

rs2

fE

t1

s4

tC

sPw

ouJ

cl

aop

an

pco

654 SVYATETS ET AL

33. Lazar HL, McDonnell M, Chipkin SR, et al: Society of Thoracicurgeons Blood Glucose Guideline Task Force. The Society of Tho-acic Surgeons practice guideline series: Blood glucose managementuring adult cardiac surgery. Report from the STS workforce onvidence based surgery. Ann Thorac Surg 87:663-669, 2009

34. Hadani M, Young W, Flamm ES: Nicardipine reduces calciumccumulation and electrolyte derangement in regional cerebral isch-mia in rats. Stroke 19:1125-1132, 1998

35. Legault C, Furberg CD, Wagenknecht LE, et al: Nimodipineeuroprotection in cardiac valve replacement: Report of an early ter-inated trial. Stroke 27:593-598, 199636. Kass IS, Cottrell JE, Chambers G: Magnesium and cobalt, not

imodipine, protect neurons against anoxic damage in the rat hip-ocampal slice. Anesthesiology 69:710-715, 198837. Zhou Y, Wang D, Du M, et al: Lidocaine prolongs the safe

uration of circulatory arrest during deep hypothermia in dogs. Can Jnaesth 45:692-698, 199838. Wang D, Wu X, Li J, et al: The effect of lidocaine on early

ostoperative cognitive dysfunction after coronary artery bypass sur-ery. Anesth Analg 95:1134-1141, 200239. Mathew JP, Mackensen GB, Phillips-Bute B, et al: Randomized,

ouble-blinded, placebo controlled study of neuroprotection with lido-aine in cardiac surgery. Stroke 40:880-888, 2009

40. Jolkkonen J, Puurunen K, Koistinaho J, et al: Neuroprotectiony the alpha2-adrenoceptor agonist, dexmedetomidine, in rat focalerebral ischemia. Eur J Pharmacol 372:31-36, 1999

41. Effects of acadesine on the incidence of myocardial infarctionnd adverse cardiac outcomes after coronary artery bypass graft sur-ery. Multicenter Study of Perioperative Ischemia (McSPI) Researchroup. Anesthesiology 83:658-673, 199542. Arrowsmith JE, Harrison MJG, Newman SP, et al: Neuropro-

ection of the brain during cardiopulmonary bypass. A randomized trialf remacemide during coronary artery bypass in 171 patients. Stroke9:2357-2562, 199843. Chong SY, Chow MY, Kang DS, et al: Deep hypothermic

irculatory arrest in adults undergoing aortic surgery: Local experience.nn Acad Med Singapore 33:289-293, 200444. Pochettino A, Cheung AT: Retrograde cerebral perfusion is

seful for deep hypothermic circulatory arrest. J Cardiothorac Vascnesth 17:764-767, 200345. Coselli JS, Buket S, Djukanovic B: Aortic arch operation: Cur-

ent treatment and results. Ann Thorac Surg 59:19-27, 199546. Coselli JS: Retrograde cerebral perfusion is an effective means

f neural support during deep hypothermic circulatory arrest. Annhorac Surg 64:908-912, 199747. Bachet J, Guilmet D, Goudot B, et al: Antegrade cerebral

erfusion with cold blood: A 13-year experience. Ann Thorac Surg7:1874-1878, 199948. Di Eusanio M, Schepens MA, Morshuis WJ, et al: Brain pro-

ection using antegrade selective cerebral perfusion: A multicentertudy. Ann Thorac Surg 76:1181-1189, 2003

49. Halstead JC, Chir MB, Meier M, et al: Optimizing selectiveerebral perfusion: Deleterious effects of high perfusion pressures.Thorac Cardiovasc Surg 135:784-791, 200850. Muller D, Fieguth H, Wimmer-Greinecker G, et al: Neurologic

utcome after surgery of the aortic arch: Comparison of deep hypo-hermic arrest, antegrade and retrograde cerebral perfusion. IJTCVS0:72-76, 200451. Coselli JS, LeMaire SA: Experience with retrograde cerebral

erfusion during proximal aortic surgery in 290 patients. J Card Surg2:322-325, 1997 (suppl)52. Safi HJ, Letsou GV, Iliopoulos DC, et al: Impact of retrograde

erebral perfusion on ascending aortic and arch aneurysm repair. Ann

horac Surg 63:1601-1607, 1997 S

53. Usui A, Oohara K, Murakami F, et al: Body temperature influ-nces regional tissue blood flow during retrograde cerebral perfusion.Thorac Cardiovasc Surg 114:440-447, 199754. Bechtel JFM, Eleftheriadis S, Sievers HH: Surgical strategies for

europrotection during ascending aortic and arch surgery. Appl Car-iopulm Pathophysiol 13:219-223, 200955. Moshkovitz Y, David TE, Caleb M, et al: Circulatory arrest

nder moderate systemic hypothermia and cold retrograde cerebralerfusion. Ann Thorac Surg 66:1179-1184, 199856. Svensson LG: Antegrade perfusion during suspended anima-

ion? J Thorac Cardiovasc Surg 124:1068-1070, 200257. Svensson LG, Hussain A, Penney DL, et al: A prospective

andomized study of neurocognitive function and S-100 protein afterntegrade or retrograde brain perfusion with hypothermic arrest forortic surgery. J Thorac Cardiovasc Surg 119:163-166, 2000

58. Camboni D, Philipp A, Schebesch KM, et al: Accuracy of coreemperature measurement in deep hypothermic circulatory arrest. In-eract Cardiovasc Thorac Surg 7:922-924, 2008

59. Marino MR, Cheng WP, Romagnoli A, et al: Temperature bias:ugular bulb venous temperature versus conventional sites. Ann Thoracurg 70:1792, 200060. Ergin MA: Hypothermic circulatory arrest, in Coselli JS,

eMaire SA (eds): Aortic Arch Surgery: Principles, Strategies andutcomes. West Sussex, UK, Willey Blackwell, 2008, pp 135-15161. Hirofumi O, Otone E, Hiroshi I, et al: The effectiveness of

egional cerebral oxygen saturation monitoring using near-infraredpectroscopy in carotid endarterectomy. J Clin Neurosci 10:79-83,00362. Orihashi K, Sueda T, Okada K, et al: Near-infrared spectroscopy

or monitoring cerebral ischemia during selective cerebral perfusionur J Cardiothorac Surg 26:907-911, 200463. Tobias JD, Russo P, Russo J: Changes in near infrared spec-

roscopy during deep hypothermic circulatory arrest. Ann Card Anesth2:17-20, 200964. Murkin JM: NIRS: A standard of care for CPB vs. an evolving

tandard for selective cerebral perfusion? J Extra Corpor Technol1:P11-P14, 200965. Denault A, Deschamps A, Murkin JM: A proposed algorithm for

he intraoperative use of cerebral near-infrared spectroscopy. Seminardiothorac Vasc Anesth 11:274-281, 200766. Schepens M, Waanders F: Monitoring the brain: Near-infrared

pectroscopy, in Coselli JS, LeMaire SA (eds): Aortic Arch Surgery:rinciples, Strategies and Outcomes. West Sussex, UK, Wiley Black-ell, 2008, pp 114-12467. Leyvi G, Bello R, Wasnick JD, et al: Assessment of cerebral

xygen balance during deep hypothermic circulatory arrest by contin-ous jugular bulb venous saturation and near-infrared spectroscopy.Cardiothorac Vasc Anesth 20:826-833, 200668. Kahn RA: EEG monitoring is required during hypothermic

irculatory arrest. Society of Cardiovascular Anesthesiologists News-etter, October 2001

69. Hagl C, Khaladj N, Karck M, et al: Hypothermic circulatoryrrest during ascending and aortic arch surgery: The theoretical impactf different cerebral perfusion techniques and other methods of cerebralrotection. Eur J Cardiothorac Surg 24:371-378, 200370. DePlessis AJ, Jonas RA, Wypij D, et al: Perioperative effects of

lpha-stat versus pH-stat strategies for deep hypothermic cardiopulmo-ary bypass in infants. J Thorac Cardiovasc Surg 114:990-1001, 199771. Sakamoto T, Kurosawa H, Shin’oka T, et al: The influence of

H strategy on cerebral and collateral circulation during hypothermicardiopulmonary bypass in cyanotic patients with heart disease: Resultsf a randomized trial and real-time monitoring. J Thorac Cardiovasc

urg 127:12-19, 2004

sb

cua

ps6

o3a

eP

at7

b

A

J

gh

iJ

pa

sc9

mJ

op

mt

it

aa

cv

sT

cC

h

dsd

eV

dm1

w6

wTp

tv

655DEEP HYPOTHERMIC CIRCULATORY ARREST

72. Muzic D, Chaney M: What’s new with alpha-stat versus pH-tat? Society of Cardiovascular Anesthesiologists Newsletter, Decem-er 200573. Shaaban-Ali M, Harmer M, Vaughan RS, et al: Changes in

erebral oxygenation during cold and warm cardiopulmonary bypasssing different blood gas strategies in patients undergoing coronaryrtery bypass grafting surgery. Acta Anesth Scand 48:837-844, 2004

74. Kiziltan HT, Baltah M, Bilen A, et al: Comparison of �-stat andH-stat cardiopulmonary bypass in relation to jugular venous oxygenaturation and cerebral glucose-oxygen utilization. Anesth Analg 96:44-650, 200375. Murkin JM, Martzke JS, Buchan AM, et al: A randomized study

f the influence of perfusion technique and pH management strategy in16 patients undergoing coronary artery bypass surgery. II. Neurologicnd cognitive outcomes. J Thorac Cardiovasc Surg 110:349-362, 1995

76. Plochl W, Cook DJ: Quantification and distribution of cerebralmboli during cardiopulmonary bypass in the swine: The impact ofaCO2. Anesthesiology 90:183-190, 199977. Dahlbacka S, Heikkinen J, Kaakinen T, et al: PH-stat versus

lpha-stat acid-base management strategy during hypothermic circula-ory arrest combined with embolic brain injury. Ann Thorac Surg9:1316-1325, 200578. Murkin J: Cerebral Autoregulation: The role of CO2 in Meta-

olic Homeostasis. Semin Cardiothorac Vasc Anesth 11:269-273, 200779. Harrington DK, Fragomeni F, Bonser RS: Cerebral perfusion.

nn Thorac Surg 83:S799-S804, 200780. Griepp RB: Cerebral protection during aortic arch surgery.

Thorac Cardiovasc Surg 121:425-427, 200181. Duebener LF, Sakamoto T, Hatsuoka S, et al: Effect of hemo-

lobin on cerebral microcirculation and tissue oxygenation during deepypothermic bypass. Circulation 104:I260-I264, 200182. Shin’oka T, Shum Tim D, Jonas RA, et al: Higher hematocrit

mproves cerebral outcome after deep hypothermic circulatory arrest.Thorac Cardiovasc Surg 112:1610-1621, 199683. Shin’oka T, Shum Tim D, Laussen PC, et al: Effects of oncotic

ressure and hematocrit on outcome after hypothermic circulatoryrrest. Ann Thorac Surg 65:155-164, 1998

84. Matalanis G, Hata M, Buxton BF: A retrospective comparativetudy of deep hypothermic circulatory arrest, retrograde and antegradeerebral perfusion in aortic arch surgery. Ann Thorac Cardiovasc Surg:174-179, 200385. Griepp RB, Ergin MA, Mc Cullough JN, et al: Use of hypother-ic circulatory arrest for cerebral protection during aortic surgery.

Card Surg 12:312-321, 1997 N

86. Mault JR, Ohtake S, Klingensmith ME, et al: Cerebral metab-lism and circulatory arrest: Effects of duration and strategies forrotection. Ann Thorac Surg 55:57-63, 199387. Lighthall GK, Cartwright CR, Haddow GR: An improvedethod for topical cerebral cooling during deep hypothermic circula-

ory arrest. J Thorac Cardiovasc Surg 120:403-404, 200088. Kaukuntla H, Harrington D, Bilkoo I, et al: Temperature mon-

toring during cardiopulmonary bypass—Do we undercool or overheathe brain? Eur J Cardiothorac Surg 26:580-585, 2004

89. The Hypothermia after Cardiac Arrest Study Group: Mild ther-peutic hypothermia to improve the neurologic outcome after cardiacrrest. N Engl J Med 346:549-556, 2002

90. Ehrlich MP, McCullough J, Wolfe D, et al: Cerebral effects ofold reperfusion after hypothermic circulatory arrest. J Thorac Cardio-asc Surg 121:923-931, 200191. McCullough JN, Zhang N, Reich DL, et al: Cerebral metabolic

uppression during hypothermic circulatory arrest in humans. Annhorac Surg 67:1895-1899, 199992. Reich DL, Uysal S, Sliwinski M, et al: Neuropsychologic out-

ome after deep hypothermic circulatory arrest in adults. J Thoracardiovasc Surg 117:156-163, 199993. Watanabe T, Washio M: Pulsatile low-flow perfusion for en-

anced cerebral protection. Ann Thorac Surg 56:1478-1481, 199394. Soeda M: Antegrade selective cerebral perfusion combined with

eep hypothermic circulatory arrest on cerebral circulation: Compari-on between pulsatile and nonpulsatile blood flows. Ann Thorac Car-iovasc Surg 13:93-101, 200795. Stier GR, Verde EW: The postoperative care of adult patients

xposed to deep hypothermic circulatory arrest. Semin Cardiothoracasc Anesth 11:77-85, 200796. Ergin MA, Uysal S, Reich DL, et al: Temporary neurological

ysfunction after deep hypothermic circulatory arrest: A clinicalarker of long-term functional deficit. Ann Thorac Surg 67:1887-1890,

99997. Svensson LG, Crawford ES, Hess KR, et al: Deep hypothermia

ith circulatory arrest. Determinants of stroke and early mortality in56 patients. J Thorac Cardiovasc Surg 106:19-31, 199398. Gandhi S, Pigula F: Strategies for aortic arch reconstruction

ithout circulatory arrest in Franco KK, Verrier ED (eds): Advancedherapy in Cardiac Surgery. Hamilton, London, BC Decker Inc, 2003,p 334-34699. Newburger JW, Jonas RA, Wernovsky G, et al: A comparison of

he perioperative neurologic effects of hypothermic circulatory arrestersus low-flow cardiopulmonary bypass in infant heart surgery.

Engl J Med 329:1057-1064, 1993