. 3
( 8)


Cannulae can be divided according to three main criteria:
• single versus two-stage (cavo-atrial);
• straight or angled; and
• metal or plastic.

Connection to the patient
This is usually achieved by right atrial (RA) cannulation. There are three basic approaches:
• Single “ A cannula is passed through the RA appendage: this route is advantageous
because it is quick and the least traumatic, but it is most sensitive to changes in position
Chapter 5: Conduct of cardiopulmonary bypass

of the heart and venous drainage can be impaired with cardiac retraction. This
technique cannot be used if the right heart is to be opened.
• Cavo-atrial “ This uses a “two-stage” cannula with a wider proximal portion, with side
holes, which lies in the RA and a narrow extension, with end and side holes, extending
into the IVC. This cannula is typically inserted through the right atrial appendage and
cannot be used if the right heart is to be opened.
• Bicaval “ Purse-string sutures are placed on the posterior“inferior RA wall and the RA
appendage to enable direct cannulation of the IVC and SVC, respectively. Tapes or
snares are passed around the vessels with the cannulae in place to ensure that the
patient™s entire venous return flows into the CPB circuit, preventing air from entering
the venous lines when opening the RA, or blood leaking past the cannulae into the RA;
this is referred to as caval occlusion or total CPB and is the technique of choice if the
right heart is to be opened.
It should be noted that the right heart may need to be vented via the pulmonary artery to
prevent RV distension due to return of blood into the RA via the coronary sinus. Occasionally
high SVC or even innominate vein cannulation may be required to facilitate the operation, for
example, resection of an RA tumor mass or during operations needing access to the SVC such
as some heart transplant or heart“lung transplant procedures (Domino heart).
In clinical practice most CABG and AVR surgery is performed with venous drainage
via cavo-atrial cannulation. This usually provides adequate drainage as long as right heart
decompression is constantly monitored and communication is maintained between the team
regarding venous return, with the necessary adjustments of cannulae position during the
operation to accommodate changes in heart position.
Air entry into the venous side of the circuit may lead to an “air lock,” causing obstruction of
venous drainage, or to systemic gaseous microemboli. The most common reason for air entry
into the system is failure to seal the site around the cannulae adequately. Care must therefore
be taken to ensure that purse-string sutures are air-tight, especially if vacuum-assisted venous
drainage is being used.

Peripheral venous cannulation
This is usually performed via the femoral or iliac veins and is used in the following instances:
• unstable patients for emergency establishment of CPB prior to sternotomy or anes-
thetic induction;
• selected redo-surgery: to provide controlled conditions during sternotomy and
exposure of the heart;
• aortic surgery;
• thoracic surgery;
• minimal access surgery; and
• extracorporeal membrane oxygenation (ECMO).
The essential requirement is for adequate venous drainage and subsequent flow rates for
CPB. With peripheral cannulation this is achieved by using as large a cannula as possible,
and passing the cannula into the RA (often using TOE guidance). Vacuum-assisted venous
drainage is advantageous under these circumstances, given the smaller cannula diameter and
increased resistance from the cannula length.
Possible complications associated with venous cannulation are listed in Table 5.3.
Chapter 5: Conduct of cardiopulmonary bypass

Table 5.3. Complications of venous cannulation

Low cardiac output due to compression of the heart during IVC purse-string placement
Damage to SVC/IVC/right pulmonary artery whilst passing tapes around cavae
Reduction in cardiac output prior to commencement of CPB when cannulae are in place
Atrial dysrhythmia
Malpositioning of cannula tip
• SVC cannula into the azygos vein
• IVC cannula into the hepatic vein
• RA cannula into LA in the presence of an atrial secundum defect
RA trauma and bleeding from cannulation sites
SVC or IVC laceration on manipulation of cannulated RA
Narrowing of cavae after decannulation and closure of purse-string suture
Low venous return during CPB
• Kinks in circuit with obstruction of line
• Reduced venous pressure “ volume or pressure related (drugs, anesthetic agents)
• Air lock
• Inadequate height of patient above CPB venous reservoir
• Inadequate size of cannula

Cardiotomy suction
During cardiac surgery “shed” blood often needs to be suctioned from the operative field
to maintain visibility for surgery and from cardiac chambers to prevent distension of the
heart. After systemic heparinization, shed blood is salvaged through designated “cardi-
otomy” suckers and vents and collected in the reservoir for re-circulation within the CPB
Cardiotomy suction is most commonly generated by use of a roller pump, requiring
repeated adjustment of pump flow rate by the perfusionist and sucker position by the surgeon
due to the degree of negative pressure that can develop at the sucker tip leading to hemolysis
of red blood cells and occasional occlusion of the sucker.
In extreme cases of hemorrhage, after heparinization with the arterial cannula in situ, the
patient can be placed on “sucker bypass”; the shed blood in the operative field provides venous
return to the CPB circuit until formal venous cannulation can be secured.

Adverse effects of cardiotomy suction
Blood suctioned from the surgical field is highly “activated” with regard to coagulation factors,
fibrinolytic mediators, leukocytes and platelets. It is a major source of hemolysis, micropar-
ticles, fat, cellular aggregates, inflammatory mediators (tumor necrosis factor alpha (TNF-±),
interleukin-6 (IL-6), C3a) and endotoxins, and a cause of platelet injury and loss. A potential
determinant of injury caused by cardiotomy suction is the amount of room air coaspirated
with the blood.
Commonly used strategies to reduce the side effects of cardiotomy suction are shown in
Table 5.4.
Chapter 5: Conduct of cardiopulmonary bypass

Table 5.4. Strategies to reduce the side effects of cardiotomy suction

Hemostasis throughout operation to minimize shed blood
Minimize aspiration of air through the cardiotomy suction by
• avoidance of high negative pressures
• slow rates of suction
• not sucking the surgical field dry
• keeping the suction tip under level of blood
Filtration of cardiotomy suction blood (leukocyte depletion)
Cell salvage blood instead of cardiotomy suction
Off-pump surgery

Venting the heart
The left side of the heart receives blood whilst on CPB from bronchial arteries and Thebesian
veins and the right heart from the coronary sinus and “leakage” around venous cannulae. As
the ventricles are unable to eject this blood during the period of arrest, a vent must be placed
to protect the heart from distension. Ventricular distension is undesirable because excessive
myocardial stretching increases myocardial oxygen demand and impairs subendocardial
On occasions, blood can return from abnormal sources. These include:
• left-sided SVC;
• patent ductus arteriosus (PDA);
• atrial septal defect/ventricular septal defect;
• anomalous venous drainage;
• aortic regurgitation; and
• systemic to pulmonary shunt.

Venting the left heart
The left ventricle needs to be vented if it is filling from any source but not ejecting. It will fill
primarily because of aortic insufficiency, or during cardioplegia administration. Venting:
• prevents distension of the ventricle;
• reduces myocardial re-warming;
• prevents ejection of air; and
• provides a bloodless surgical field.
Surgical inspection and palpation of the LV to monitor the degree of distension is
crucial on commencing CPB, during aortic cross-clamping and during initial administration
of cardioplegia. The use of a left atrial pressure monitoring line and pulmonary artery (PA)
catheter can help detect moderate LV distension, which is sometimes a subtle finding.

Venting the right heart
The venous cannulae effectively vent the right side of the heart, keeping it empty of
blood except for any “leakage” past the cannula; this can be minimized by using bicaval
Chapter 5: Conduct of cardiopulmonary bypass

cannulation and caval snares. When antegrade cardioplegia is administered, releasing the
caval snare will permit venting of cardioplegia solution returning via the coronary sinus to
the right heart.
Placement of a pulmonary arterial vent will keep the right ventricle empty of fluid. Persist-
ent left SVC requires additional drainage of the coronary sinus or RA.

Venting methods
Venting can be achieved via:
• the aortic root cardioplegia cannula “ this method does not allow venting during
cardioplegia administration;
• the right superior pulmonary vein “ a vent is passed into the left atrium and through
the mitral valve into the LV;
• the left ventricular apex; and
• the pulmonary artery “ this may not be effective at venting the LV when there is aortic
regurgitation with a competent mitral valve.
It must be remembered that venting the heart is not without complications. These can
be immediate or delayed. “Steal” of systemic perfusion may occur if excessive venting of the
heart is employed in the presence of aortic regurgitation. Systemic air embolism may occur
when the vent is inserted or removed. Bleeding can occur from the vent site, particularly if an
LV apical vent is used. Later complications of venting include stenosis of the pulmonary vein
or pulmonary artery, or aneurysm of the LV apex, depending on the vent site used.

General management of CPB
Before commencing CPB the perfusionists must have completed a series of “checks” as
detailed in Chapter 2.

Transition of patient onto CPB
The general sequence of events entailed in commencing CPB is described below and sum-
marized at the end of this chapter in Appendix 5A:
• The patient must be systemically heparinized with confirmation of adequate
anticoagulation (ACT >400 s). ACT targets vary among institutions and are the subject
of much practical and academic debate. Heparin resistance must be considered if a
therapeutic ACT cannot be achieved despite additional systemic boluses of heparin.
Antithrombin III (AT-III) deficiency can be treated by giving AT-III-rich blood
products such as FFP or AT-III concentrate. Repeat ACT should result in an elevated
reading without the need for further heparin.
• Recirculation of CPB prime through the circuit to ensure that perfusate is warm and
lines are air free.
• Division of the arterial and venous lines after clamps are applied at the patient and
pump ends.
• The arterial line is connected to the aortic cannula, enabling rapid transfusion, if
required, directly from the pump. During connection to the arterial line, the arterial
cannula is allowed to bleed out slowly from the patient and the CPB prime is
simultaneously advanced up the arterial line to ensure an air-free connection.
Chapter 5: Conduct of cardiopulmonary bypass

• Release of the aortic line clamp, after which the pressure “swing” is confirmed to
indicate correct positioning of the aortic cannula.
• Retrograde autologous priming may be used to reduce the effect of hemodilution due
to large prime volumes in the CPB circuit. The cardiotomy reservoir is primed to a
minimal level and the patient™s arterial blood is used to fill the circuit to a “safe” level
after placement of the arterial cannula (this is approximately 400 ml). Hemodynamic
consequences of this technique need to be considered for each patient. This technique
is discussed in more detail in Chapter 3.
• The venous cannula is connected to the venous line.
• CPB is initiated, after instructions from the surgeon, by the perfusionist releasing the
arterial line clamp and slowly transfusing the patient with the CPB prime. Arterial flow
should be unobstructed and with an initial line pressure of less than 100 mmHg. The
perfusionist must confirm that the oxygenator gases and CPB safety alarms are
switched on prior to CPB.
• The venous clamp is gradually released after confirmation that the arterial line is
unobstructed; the patient™s venous blood is then diverted into the circuit. The right
heart should decompress with a fall in CVP to less than 5 mmHg.
• A period of 1“2 minutes of transition occurs whilst the perfusionist gradually increases the
rate of arterial flow and venous return to the heart is reduced. The arterial pressure changes
to a non-pulsatile waveform. Pulsatility whilst on CPB indicates aortic valve insufficiency,
inadequate venous drainage or excessive bronchial venous return into the heart.
• Most CPB systems generate non-pulsatile flow, but some do have computerized
configurations to allow for pulsatile flow generation. Alternatively, pulsatile flow can be
generated via an intra-aortic balloon pump, if in situ. There are many reported benefits
of pulsatile flow such as increased renal, cerebral and myocardial perfusion together
with a reduction of the stress response to CPB. However, in clinical studies, the
conclusions are equivocal.
• Cooling of the patient, if required by the surgeon, is commenced once the patient is on
full flow and adequate decompression of the heart is confirmed.

Recommended flow rates for CPB
The primary requirement for CPB is to provide a systemic O2 delivery (DO2) that is sufficient
to meet systemic O2 demand (VO2). In contrast to the intact native circulation, DO2 is not
controlled by reflex mechanisms, but by the perfusionist. During CPB whole body DO2 is a
function of pump flow and arterial oxygen content, the latter being primarily determined by
the HCT. The major determinants of VO2 are temperature and level of anesthesia.
Oxygen consumption (VO2) can be calculated using the Fick equation:

VO2 = Q(C(a-v) )O2
VO2 = minute oxygen consumption (ml/minute)
Q = cardiac output (l/minute)
(C(a-v) )O2 = 1.34 — Hb + P(a-v)O2

where 1.34 is the hemoglobin oxygen content at 100% saturation (ml/g), Hb is the hemo-
globin concentration (g/l) and P(a-v)O2 is the arterio-venous oxygen partial pressure difference
Chapter 5: Conduct of cardiopulmonary bypass

Table 5.5. Recommended flow rates (l/minute) for different surface areas and flow index

Body surface area (m2) Flow index 1.8 l/minute/m2 Flow index 2.2 l/minute/m2
1.60 2.88 3.52
1.80 3.24 3.96
2.00 3.60 4.40
2.20 3.96 4.84

Effective blood flow is that which actually results in maintenance of near-physiological tis-
sue perfusion. Effective perfusion is reduced by anatomical shunting of arterial blood around
the capillary bed to the venous circulation, e.g., via bronchial or pulmonary collaterals, and by
the physiological shunt created by blood suctioned from the surgical field.
Indices of adequate total perfusion include pH, lactate and SvO2 (hemoglobin oxygen satura-
tion in venous blood). A low SvO2 during CPB indicates an imbalance between DO2 and VO2
and requires a change in perfusion conditions. It can reflect insufficient pump flow, HCT, hemo-
globin oxygen saturation, inadequate anesthesia or increasing temperature. All of these param-
eters should be optimized to ensure effective blood flow and consequent perfusion of organs.
In adults at normothermia, clinical and experimental data support a minimum flow index
of 1.8 l/minute/m2. Kirklin and Barratt-Boyes recommended a flow index of 2.2 l/minute/m2
for adults at a temperature of 28°C or above. The patient™s body surface area (m2) is worked
out from a normogram plotting height in meters and weight in kilograms. Patients with
a body surface area greater than 2 m2 should have the flow maintained at 1.8“2.2 l/minute/m2 to
avoid excessively high flows through the machine leading to hemolysis.
Table 5.5 shows recommended flow rates for different surface areas.
Flow rates are reduced at lower body temperatures as VO2 also decreases (see Appendix 5A).

As the CPB circuit is primed with crystalloid or colloid, hemodilution of the patient inevitably
results. The degree of hemodilution caused by CPB can be calculated before initiating bypass
so that the prime solution can be adjusted to incorporate packed red blood cells if unaccept-
able levels of anemia are anticipated.
preop HCT — PBV
PBV + CPB prime volume
where PBV = patient™s blood volume (l) and CPB prime volume = extracorporal prime
volume (l).
Benefits of hemodilution include reduced blood viscosity and an increase in microvas-
cular blood flow, but these effects are partially counterbalanced by the reduction in oncotic
pressure, which may promote tissue edema.

Mean arterial blood pressure (MAP)
An acceptable MAP on CPB is that which provides adequate tissue perfusion. Adequate tissue
perfusion is, however, also influenced by the pump flow rate and the core body temperature.
MAP is determined by flow rate and arteriolar resistance. In general, higher pressures should
be maintained in the presence of known cerebrovascular disease, in particular carotid steno-
sis, renal dysfunction, coronary disease or left ventricular hypertrophy.
Chapter 5: Conduct of cardiopulmonary bypass

On commencement of CPB there is a transient drop in systemic pressure. This is due
to vasodilatation associated with the sudden decrease in blood viscosity resulting from
hemodilution by the CPB prime solution and, secondarily, from the systemic inflammatory
response (SIRS) associated with CPB. However, as CPB continues there is a gradual increase
in perfusion pressure due to increasing vascular resistance. This is a result of equilibration of
fluid between the vascular and tissue “compartments,” hemoconcentration from diuresis, the
increase in blood viscosity seen with hypothermia and the progressive increase in circulating
levels of catecholamines and renin as part of the stress response to CPB.
It is important to emphasize that manipulation of MAP alone is not sufficient to guarantee
adequate organ perfusion. Neither a low MAP with a high flow nor a high MAP with a low
flow are sufficient in themselves. Whole body DO2 must firstly be optimized and secondly,
vascular resistance altered to bring the MAP into the autoregulatory range for critical organ
beds, with due consideration to underlying pathophysiology.

Pulmonary artery and left atrial pressure
On CPB the PA and LA pressures should be close to zero. PA or LA pressure monitoring is
useful during CPB to assess left ventricular distension, in particular in cases where increase
in blood flow back to the left heart is expected (cyanotic heart disease, large bronchial flow in
chronic lung disease or aortic regurgitation). Care must be taken with PA catheters to ensure
that migration of the catheter tip does not occur, leading to “wedging” and subsequent PA
rupture or infarction of the lung.

Central venous pressure
On CPB, CVP is expected to be close to zero and no more than in single digits. An increase in
CVP indicates impaired venous drainage to the reservoir. The causes of an increase in CVP are
inadequate cannula size, obstruction to the line or cannula tip and insufficient height differ-
ence between the patient and the reservoir to enable gravity siphon drainage. The consequence
of an increase in CVP during bypass is to reduce effective perfusion of critical organs with
resultant edema. The liver is particularly sensitive to reduced flow as nearly three-quarters
of hepatic blood flow occurs at near venous pressure. If a persistently high CVP, uncorrected
by attention to the factors mentioned above, is noted during CPB the patient™s head and eyes
should be closely observed for signs of engorgement and consideration given to altering the
venous cannulation to improve drainage.

The ECG must be recorded throughout CPB to ensure that it remains isoelectric during
cardioplegic arrest. Following removal of the aortic clamp and resumption of myocardial
activity persistent ST segment changes may be related to ischemia resulting from inadequate
re-vascularization, coronary ostial obstruction, e.g., by an incorrectly seated aortic valve
prosthesis, or air/particulate embolization. Additionally, the ECG is useful in guiding the
postoperative management of epicardial pacing.

The principal reason for hypothermic CPB is to protect the heart and other organs by reduc-
ing metabolic rate and thus oxygen requirements. In the myocardium, hypothermia sus-
tains intracellular reserves of high-energy phosphates and preserves higher intracellular pH
Chapter 5: Conduct of cardiopulmonary bypass

and electrochemical neutrality. Myocardial cooling can be achieved with cold cardioplegia,
pouring cold topical solution on the heart and cooling jackets, as well as by systemic hypo-
thermia. Systemic hypothermia is not uniform due to different blood flow to different vascular
beds. High blood flow rates and slow cooling ensures less variation in systemic hypothermia.
Temperature should be measured at multiple sites and the advantages and limitations of each
site needs to be recognized. During cardiac surgery temperature can be measured in the fol-
lowing locations: nasopharynx, tympanic membrane, pulmonary artery, bladder or rectum,
arterial inflow, water entering heat exchanger and venous return.
Nasopharyngeal temperature probes underestimate, but approximate to brain tempera-
ture, with the mixed venous temperature on the CPB circuit being an approximation of average
body temperature. Bladder and rectal temperatures give an indication of core body tempera-
tures, but these can be erroneous due to interference from varying urine production and fecal
matter, respectively. These low blood flow sites tend to underestimate temperature so are par-
ticularly valuable following deeper levels of hypothermia. On re-warming the aim is to achieve
uniform normothermia. To avoid rebound hypothermia after cessation of CPB, which occurs
if too great a temperature gradient is allowed to develop between peripheral and core tempera-
tures, vasodilators can be used to promote more uniform re-warming by distributing greater
blood flow, and therefore heat, from the core to peripheries. The process of re-warming must be
controlled to avoid rapid changes in temperature, or excessive blood temperatures, which can
result in micro-bubble formation due to the reduced solubility of gases in blood as the tempera-
ture increases, denaturing of plasma proteins, hemolysis and cerebral injury. As a general guide
for every 1°C drop in temperature there is an associated 7% drop in oxygen demand, i.e., a 7°C
reduction in temperature results in a 50% drop in oxygen demand (see Table 5.6).
At less than <15°C oxygen is too tightly bound to hemoglobin and is therefore unavailable
to tissues. In addition, the viscosity of the blood can be too high for effective flow through the
CPB circuit.

Urine volume
Urine volume on CPB is monitored as an indicator of renal perfusion. Indications for diuretic
use during CPB include hyperkalemia, hemoglobinuria and hemodilution. Furosemide is
used for treatment of hyperkalemia and mannitol is used to generate alkaline urine to treat

Transesophageal echocardiography (TOE)
TOE is applied increasingly as a routine part of surgery when intracardiac cavities are opened.
It is a useful tool to assess adequacy of de-airing of the heart. In addition, TOE can be used to

Table 5.6. Hypothermia: temperature ranges and indications for use

Hypothermia Temperature (°C) Use
Tepid 33“35 Good for short operations, healthy
patients with higher HCTs
Mild 31“32 Protection of beating heart and
neurological systems
Moderate 25“30 Protection of non-beating heart and
neurological systems
Deep 15“20 DHCA for typically 40“60 minutes

Chapter 5: Conduct of cardiopulmonary bypass

assess intracardiac structures (valves, prostheses, septal walls, left and right ventricular out-
flow tracts) and regional wall motility.

Laboratory investigations
This is discussed in detail in Chapter 6.
Minimal monitoring during CPB requires measurement of PO2, PCO2, base excess, hemo-
globin, HCT, pH, potassium, glucose and coagulation status using ACT.

Termination of CPB
Table 5.7 provides a checklist of the basic conditions that need to be fulfilled before wean-
ing can be attempted. Terminating CPB is a gradual process with constant communication
between surgeon, anesthetist and perfusionist. The first step is for the perfusionist to restore
the blood volume to the heart by gradual occlusion of the venous return. Th e patient is par-
tially supported by the CPB machine with blood passing through both the heart and the lungs.
The heart begins to eject blood when a critical volume is reached. The perfusionist continues
to return blood from the venous reservoir to the patient whilst continuing to occlude the
venous line until the patient is weaned from CPB. Termination of CPB is achieved by com-
plete occlusion of the arterial and venous lines.
Transfusion of blood to the patient is still possible through the arterial cannula follow-
ing cessation of CPB. The venous cannula is removed when the patient is stable and the
process of reversing heparin with protamine is due to commence. Some surgeons leave
the venous purse-string suture untied but snared to enable rapid re-insertion of a cannula
for emergent return to CPB if required. Prior to protamine administration cardiotomy
suction is stopped to avoid clotting within the bypass circuit. The protamine should be
administered slowly due to its propensity for causing systemic vasodilatation and pulmo-
nary vasoconstriction. Transfusion of residual blood from the pump is usually required to
support cardiac filling during protamine administration; generally boluses of 100 ml are
given, titrated against MAP and CVP, PA or LA pressures and direct observation of the
heart. The aortic cannula is typically removed when protamine admistration is completed,
the patient is stable and there is no further requirement for transfusion of residual blood
via the CPB machine. The two purse-string sutures on the aorta are tied to secure the can-
nulation site.

Table 5.7. Checklist before weaning from CPB

Patient position on operating table is neutral
Operation completed and vent sites closed
Hemostasis secured
Heart de-aired (confirmed with TOE if available)
Ventilation of lungs recommenced and adequate
Acceptable Hb/HCT, potassium, glucose, and acid“base status on arterial blood gas analysis
Acceptable core temperature achieved
Heart rhythm and rate appropriate
Parameters for initial filling pressure when off CPB are determined
Inotropic support prepared if necessary

Chapter 5: Conduct of cardiopulmonary bypass

The remainder of the blood in the bypass circuit can be retained for transfusion by the
anesthetist directly or it can be processed through a cell-salvage device to maximize the red
cell concentration of this “pump blood.” After transfusion a further dose of protamine may be
administered to counteract the heparin in the pump blood.
Transition from CPB to physiological circulation is more often than not an unevent-
ful process. In some circumstances, particularly when operating on patient™s with severely
impaired ventricular function, or if there has been a long ischemic period during the proce-
dure, weaning from CPB may require measures to be taken to support the circulation. Such
measures are discussed in Chapter 8.

Appendix 5A: Protocol for the conduct of “routine” CPB
(Adapted, with permission, from London Perfusion Science Protocols.)

5A.1: Connection of the circuitry to the patient
• The surgeon will ask for the CPB lines to be divided
• The pump flow is slowly reduced and the venous line clamped, followed by the arterial
line, beyond the recirculating Y-connector
• The surgeon will cannulate the aorta or peripheral arterial vessel
• If required, the arterial pump is continuously turned to assist an air free connection
• When the line is free of air the surgeon will connect the arterial line and confirm that
the connection is satisfactory
• The clamp from the arterial line is removed and repositioned behind the recirculating
• The swing on the “Tycos” gauge is checked
• If required, 50 ml blood is transfused to determine the adequacy of the swing
• Perfusionist should state there is a “good swing” if the gauge swings freely and reflects
the patient™s blood pressure
• If the perfusionist has any doubts about the cannulation, he must inform the surgeon
immediately, continuing to voice his misgivings until he is confident that the cannula is
satisfactorily placed
• Surgeon will cannulate the venous circulation (via RA, IVC and SVC or peripherally)
• Be prepared to use the pump suckers to deal with any blood loss
• Be prepared to transfuse the patient to replace this lost blood volume

5A.2: Initiating bypass
• If the gases have not yet been switched on, they are now correctly set according to the
patient™s rated flow
• The perfusionist must now clearly inform the medical staff that he/she is “going onto
• The clamp on the arterial line is removed and the pump is slowly turned at first,
gradually increasing the rpm
• When going onto bypass with a centrifugal pump, forward pressure must be generated
before the line clamp is removed. The drive motor is therefore turned on whilst the
aortic line is still clamped, in order to generate sufficient forward pressure, to exceed
the patient™s arterial pressure; above 2000 rpm is usual
Chapter 5: Conduct of cardiopulmonary bypass

• The perfusionist must monitor the pressure on the line pressure (electronic or Tycos)
during this stage, looking for any sign of obstruction: at the same time monitoring the
venous and arterial pressures and, of course, monitoring the blood level in the venous
reservoir, as the pump speed is increased
• Having raised the pressure on the venous side, the venous clamp is removed “ more
quickly if air has been left in the venous line “ until this air has been removed.
Perfusionists should then control the venous pressure with their clamp until they
have achieved full rated flow for the patient
• The anesthetist should be informed when full flow has been achieved, so that
ventilation may be discontinued
• Any difficulty in achieving a full venous return should be reported immediately to the
surgeons, so that they may make any adjustment to the venous cannulation as may be
necessary. Venous air should also be reported to the surgeon. It is important that an
optimum venous return can be obtained at this stage
• The perfusionist must monitor the ECG at this stage, so that arrhythmias, particularly
ventricular fibrillation (VF), may be noted early and action taken to prevent cardiac
• Once the aorta has been clamped, the required temperature has been achieved,
cardioplegia has been administered, if required, and a steady state of perfusion has
been attained, the first sample for blood gases, electrolytes, ACT, Hb/HCT and glucose
is taken

5A.3: Patient flows
The flow required to meet a patient™s metabolic requirements may need to be modified in
certain circumstances (such as the presence of carotid disease). Using the patient™s height
and weight, the patient™s surface area is obtained from a standard nomogram, and hence
flows calculated for differing levels of metabolic requirement:

Flow index
Hypothermia Temperature (°C)
Normothermia 34“37 2.4
Moderate hypothermia 32“34 2.2
Hypothermia 28“32 1.8“2.0
Profound hypothermia <28 1.6

As a general rule, flows should be reduced with temperature (as metabolic requirement
diminishes) and vice versa. Whilst individual cases may require special consideration, it is
important to note the following:
Hypothermia is used as a technique in order that flows may be safely reduced. Too high
a flow at a reduced temperature may:
• Cause blood damage
• Impede the surgery by flooding the field
• Cause an excessive rise in venous pressure
• Cause warming of the heart when cardioplegia has been used
• Cause underperfusion. Flows should correspond to temperature. Too low a flow during
re-warming or at normothermia may lead to serious underperfusion
Chapter 5: Conduct of cardiopulmonary bypass

5A.4: The re-warming phase
The re-warming phase begins only after consultation with the surgeon. On re-warming,
appropriate adjustments to gas flows and to blood flows must be made. This is a period
during which a rapid drop in SVO2 may be experienced. A sample for all parameters should
be taken during the mid-warming phase, in order to give sufficient time for any corrective
action to be taken before coming off bypass. Final samples should be taken once a core
temperature >35°C has been attained.

5A4.1: Re-warming
• The patient should be re-warmed using the arterial blood temperature and patient core
temperature as guides to the rate and extent of re-warming
• The target arterial blood temperature is between 37.5 and 38°C. The upper limit should
not be exceeded
• A gradient of 10°C between the water temperature in the heater“chiller unit and the
arterial blood should not be exceeded
• Re-warming the patient to 37°C (nasopharyngeal) is usually a maximum, although
surgeons vary in this regard
• Appropriate adjustments to gas flows and to blood flows must be made
• The rate of re-warming should be such as to allow time for distribution of heat between
core and peripheral tissues, using vasodilators, if needed, to enhance peripheral blood
flow and thus heat distribution
• Post-CPB an “after drop” in core temperature occurs as heat is redistributed from core
to peripheral tissues; this after drop can be lessened if adequate time is allowed for
thorough re-warming

cardiopulmonary bypass: a safe
Suggested Further Reading and effective means of decreasing
• Abel RM, Buckley MJ, Austen WG, Barnett hemodilution and transfusion
GO, Beck CH Jr, Fischer JE. Etiology, requirements. J Thorac Cardiovasc Surg
incidence, and prognosis of renal failure 1998; 115(2): 426“39.
following cardiac operations. Results of a
• Nuetzle, Bailey CP. New method for
prospective analysis of 500 consecutive
systemic arterial perfusion in
patients. J Thorac Cardiovasc Surg 1976;
extracorporeal circulation. J Thorac Surg
71(3): 323“33.
1959; 37(6): 707“10 (no abstract available).
• Bennett EV Jr, Fewel JG, Grover FL, Trinkle
• Rosengart TK, DeBois W, O™Hara M, et al.
JK. Myocardial preservation: effect of
Deairing of the venous drainage in standard
venous drainage. Ann Thorac Surg 1983;
extracorporeal circulation results in a
36(2): 132“42.
profound reduction of arterial micro
• Hartman GS, Isom OW, Krieger KH. bubbles. Thorac Cardiovasc Surg 2006;
Retrograde autologous priming for 54(1): 39“41.

Metabolic management during

cardiopulmonary bypass
6 Kevin Collins and G. Burkhard Mackensen

The key to metabolic management during cardiopulmonary bypass (CPB) is the mainten-
ance of adequate blood flow and oxygen delivery to the body™s tissues. Utilizing the CPB
machine, the perfusionist provides the optimum conditions necessary for operations on the
heart, lungs or major vessels, while supporting the patient™s physiological and metabolic

The perfect perfusion to me¦ is to be allowed to perform the necessary repair, however long
that takes and yet leaving my patients looking like they™ve never been on bypass.
Dr. Norman Shumway, Stanford University

CPB-induced perturbations of patient metabolism
and corrective interventions
CPB induces a unique set of physiological disturbances. The principal causes of metabolic
derangement include:
• fluid priming of the CPB circuit;
• organ hypoperfusion; and
• changes in body temperature.
The causes, management and monitoring of metabolic parameters during CPB are dis-
cussed in this chapter.

CPB primes, hemodilution, autologous priming
and hemofiltration
CPB circuit primes
All priming fluids cause hemodilution, which leads to a fall in the hematocrit (HCT), altera-
tions in the volume of distribution of electrolytes and fluid shifts between the vascular and
intercellular compartments. Every attempt should be made to minimize the volume of the
CPB circuit. Use of small-diameter tubing and cannulae in the circuit, minimizing the length
of circuit tubing, partially priming with autologous blood and using vacuum-assisted venous
drainage all provide easy methods of reducing priming volume.
In the earliest days of cardiac surgery and CPB in the late 1950s, the prime was constituted
to provide near normal HCT levels. However, with the advent of the use of hypothermia in
the 1960s, intentional hemodilution became standard practice. Hemodilution is principally
70 Cardiopulmonary Bypass, ed. S. Ghosh, F. Falter and D. J. Cook. Published by Cambridge University Press.
© Cambridge University Press 2009.
Chapter 6: Metabolic management

Table 6.1. Expected HCT on CPB

HCT on CPB = red cell volume/system volume
Red cell volume = patient blood volume — pre-CPB HCT
System volume = patient blood volume + prime volume
Blood volume:
Adult “ male: 70 ml/kg, female: 60 ml/kg
Child “ 1“10 years: 80 ml/kg, 3“6 months: 85 ml/kg, 0“3 months: 90 ml/kg

the result of the need for fluid priming of the CPB circuit, but also arises from the infusion of
fluids during surgery and the administration of cardioplegia solution (CPS). Some degree of
hemodilution is considered to be beneficial as:
• the reduction in blood viscosity minimizes CPB circuit sheer stresses upon blood,
thereby lowering hemolysis rates; and
• reduced blood viscosity improves blood flow through capillary networks.
Utilizing weight-adjusted formulas, the desired HCT on CPB can be calculated as out-
lined in Table 6.1.
Further rationale for the use of hemodilution include reducing bank blood usage and
the associated risks of transfusion, as well as respecting the wishes of patients not wanting to
receive blood transfusions (e.g., Jehovah™s Witness).
Hemodilution affects the concentration of plasma proteins. These colloids exert an oncotic
pressure, holding water in the vascular compartment and so preventing the accumulation of
water in interstitial spaces. Plasma proteins also bind a high proportion of drugs and electro-
lytes, maintaining a balance between their unbound, ionized state and their protein-bound
state. Serum albumin (3.5“5.5 g/dl) constitutes 50“70% of the total protein with globulins
(2“3.6 g/dl) comprising the bulk of the remainder.
Starches, modified animal colloids or human albumin may be added to CPB primes to
increase their effective oncotic pressure. The composition of CPB primes is discussed in detail
in Chapter 3.

Autologous priming
Autologous priming (AP) utilizes the patient™s blood to re-prime the CPB circuit upon initia-
tion of CPB. Normal (antegrade) blood flow (Q) through the CPB circuit displaces the circuit
prime with the patient™s venous blood while diverting the crystalloid into a sterile bag. During
the 10- to 20-second period required for AP, the patient is essentially being exsanguinated
and the anesthetist must administer vasoconstrictors as required to maintain blood pres-
sure. Alternatively, partial priming with autologous arterial blood can be achieved by retro-
grade drainage of 100“400 ml of blood via the arterial cannula into the cardiotomy reservoir.
Depending on circuitry type, AP can significantly reduce the amount of crystalloid prim-
ing volume (e.g., from 1500“1800 ml to ˜1100“1400 ml). AP can dramatically decrease the
extent of hypotension, attributed to rapid hemodilution, commonly seen following the initia-
tion of CPB. AP allows for higher HCT levels with slightly higher viscosities at warmer CPB
temperatures (32“35°C) and still appears to avoid hemolysis from CPB circuit sheer stresses.
AP also aids “normalization” of vascular oncotic pressures, thus decreasing fluid shifts and
“third spacing.”
Chapter 6: Metabolic management

Table 6.2. Hemodynamic calculations for CPB

BSA = √ (kg — cm/3600)
CPB flow (Q) = CI — BSA
SVR = (MAP “ CVP/CO) — 80
BSA “ body surface area; CI “ cardiac index; CO “ cardiac output; CPB “ cardiopulmonary bypass; CVP “ central
venous pressure; Q “ blood flow; SVR “ systemic vascular resistance; MAP “ mean arterial pressure.

Table 6.3. O2 calculations for CPB

VO2 = (SaO2 “ SvO2) (1.34) (Hb) + (PaO2 “ PvO2) (0.003)
O2 capacity = (1.34) (Hb) + (0.003) (PO2)
O2 content = (1.34) (Hb) (SaO2 or SvO2) + (0.003) (PO2)
CPB O2 consumption = (aO2 “ vO2) (Q l/minute) (10)
CPB O2 transfer = [(SaO2 “ SvO2) (1.34) (Hb) (Q ml/minute)] /100
aO2 content “ arterial oxygen content ; CPB “ cardiopulmonary bypass; Hb “ hemoglobin; PaO2 “ oxygen arterial
partial pressure; PvO2- oxygen venous partial pressure; Q “ blood flow; SaO2 “ arterial oxygen saturation; SvO2 “
mixed venous oxygen saturation; vO2 content “ venous oxygen content.

CPB flow rates
The perfusionist calculates a CPB blood flow (Q) utilizing the patient™s body surface area
(BSA) and cardiac index (CI) (see Table 6.2). Insufficient flow can result in inadequate tissue
perfusion. Metabolic acidosis during CPB is almost always the result of hypoperfusion lead-
ing to oxygen delivery inadequate to meet metabolic demands for aerobic respiration. Oxygen
consumption (VO2) is thus a major determinant of CPB flow requirements (see Table 6.3).

CPB-related hypoperfusion may be intentional or unintentional. The most common inten-
tional causes of transient periods of hypoperfusion are induced by the cardiac surgeon.
Manipulation of the heart may impede venous blood return to the CPB circuit necessitating a
reduction in flow. Frequently, the surgeon requests the perfusionist to reduce flow to permit
safe application or removal of the aortic cross-clamp, decrease surgical bleeding or to empty
and decompress the heart.
Reduction of pump flow, for unintentional reasons, is nearly always caused by poor venous
return to the circuit, usually due to a venous cannula that has been advanced too far or surgi-
cal distortion of the vena cavae and heart, although improper arterial cannulation, movement
of the cannulae after placement or an incorrectly occluded arterial roller pump head some-
times may occur.

Temperature and hypothermia
The temperature of the patient whilst on CPB is one of the most profound determinants of the
requirements for perfusion. Systemic O2 consumption, VO2, is reduced by approximately 50%
for every 7°C reduction in core temperature below normothermia (30°C = 50%, 23°C = 25%,
16°C = 12.5% metabolic demand of the same organ at 37°C; see table 6.4). As such, relatively
Chapter 6: Metabolic management

Table 6.4. Classification of hypothermia

Mild: 36“34°C
Moderate: 33“28°C
Severe: 27“22°C
Deep: <21°C

Figure 6.1 Oxyhemoglobin
dissociation curve.

small decreases in temperature markedly reduce the requirements for systemic O2 delivery,
making moderate reductions in pump flow or HCT tolerable, such that DO2 remains suf-
ficient to meet VO2.
Left and right shifts in the oxygen“hemoglobin dissociation curve also occur with tem-
perature changes during CPB (see Figure 6.1). At lower temperatures hemoglobin has a
greater affinity for binding oxygen, consequently oxygen is also released less readily and the
dissociation curve is shifted to the left. At higher temperatures the converse is true and the
curve is shifted to the right.
Pump flow rate must be adjusted with due consideration to temperature if the metabolic
demands for oxygen are to be matched by delivery. Typical flow rates over a range of tempera-
tures are shown in the Appendix 5A.

Deep hypothermic circulatory arrest (DHCA)
Deep hypothermic circulatory arrest is discussed in detail in Chapter 10, but is briefly men-
tioned here for completeness. Certain cardiac procedures require DHCA, rather than just
conventional mild to moderate hypothermia, usually because the aorta cannot be cross-
clamped or total absence of blood flow is required to enable surgical access. DHCA is used
Chapter 6: Metabolic management

to dramatically lower the body™s metabolic demand while protecting organs, particularly the
brain, during a period in which perfusion is suspended. This technique utilizes profound
hypothermia, with or without the use of aortic cross-clamping and delivery of cardiople-
gia, to facilitate surgery to the left ventricular outflow tract, aortic valve, ascending aorta or
great vessels. Pediatric palliative and corrective surgical procedures also frequently neces-
sitate periods of DHCA. Procedurally dictated, intermittent “low-flow” (5“15 ml/kg/minute)
states may be employed during DHCA to deliver oxygenated blood to the brain via antegrade
(ACP) and retrograde cerebral perfusion (RCP). During ACP in adult patients, mean arterial
pressure (MAP) should be ¤65 mmHg, and during RCP the central venous pressure (CVP)
should be ¤25 mmHg.

pH, acid“base, blood gases and electrolytes
pH and acid“base
The normal pH of arterial blood is 7.4 (± 0.05). Bicarbonate and non-bicarbonate systems
play important roles in buffering pH changes.

Bicarbonate system
The bicarbonate buffer system (carbonic acid H2CO3 and bicarbonate HCO3“) is considered
to be the most important mechanism for physiological regulation of pH. It possesses approxi-
mately 53% of the total buffering capacity of body fluids. Exogenous sodium bicarbonate is
easily administered during CPB. It should be noted that bicarbonate™s molecular weight is
small enough to allow its passage across the semipermeable fibers of the hemofilter and may
thus be removed with the effluent product or “plasma water waste” if hemofiltration is used
during CPB. The simple formula [(body weight (kg) — 0.3)/2] — base deficit = mmol NaHCO3“
needed to yield base excess equal to 0 is often used when treating persistent acidosis.

Non-bicarbonate buffers
• Inorganic phosphate buffers are important in regulating pH in the intracellular and
renal tubular fluids. Inorganic phosphates, like bicarbonate, are removed during
• Plasma proteins possess significant buffering capacity because of the ionic nature of
their amino acid structure and because of their high plasma concentrations. Plasma
proteins are not removed during hemofiltration because of their larger molecular size.
• Hemoglobin and oxyhemoglobin play a major role in buffering hydrogen ions at the
tissue level. Considered the most important of the non-bicarbonate pH buffers,
hemoglobin is not removed during hemofiltration because of the size of the red blood

Metabolic acidosis and alkalosis
Metabolic acidosis is usually due to systemic O2 delivery (DO2) during bypass not meeting
systemic O2 demand (VO2). The options to address this are to increase pump flow or HCT,
thereby increasing DO2, or to reduce VO2 by decreasing temperature or possibly by increasing
depth of anesthesia. Failing this, administration of sodium bicarbonate, or the use of hemofil-
tration (ultrafiltration) may correct the acidosis.
Chapter 6: Metabolic management

CPB-related metabolic alkalosis is usually due to a reduction in serum potassium levels
(e.g., due to increased urine output or hemofiltration) and is best treated by titrated adminis-
tration of potassium chloride.

Respiratory acidosis and alkalosis
Respiratory acidosis is the result of insufficient removal of CO2 from the patient™s blood by
the membrane oxygenator. Increasing the sweep gas rate through the membrane oxygenator
will facilitate the transfer or elimination of excess CO2 from the patient™s blood. Conversely,
respiratory alkalosis is the result of excessive CO2 removal.

Alpha-stat and pH-stat strategies for blood gas management
The optimal pH management strategy during hypothermic cardiopulmonary bypass is as yet
undetermined. The two main strategies utilized clinically, alpha-stat and pH-stat, differ in their
approach to the acid“base alterations that occur with hypothermia. As blood temperature
falls, gas solubility rises and the partial pressure of carbon dioxide decreases (PCO2 decreases
4.4% for every °C drop in temperature). With alpha-stat management, arterial gas samples
are not corrected for sample temperature and the resulting alkalosis remains untreated dur-
ing cooling; with pH-stat management, arterial blood gas samples are temperature corrected
and carbon dioxide is added to the gas inflow of the CPB circuit so that the PCO2, and hence
pH, is corrected to the same levels as during normothermia. The advocates of alpha-stat point
to potential benefits in terms of the function of intracellular enzyme systems and the advan-
tage of preserving cerebral autoregulation. Proponents of pH-stat, which results in cerebral
vasodilation, cite as advantages higher levels of oxygen delivery to the brain and enhanced
distribution of blood flow. However, the higher cerebral blood flows associated with pH-stat
also have the potential to carry more gaseous or particulate emboli to the brain.
Alpha-stat management is based on the concept that the dissociation constant, pK, of the
histidine imidazole group changes with temperature in a manner nearly identical to physio-
logical blood buffers. Hence, the ionization state (±) of this group stays the same, irrespective
of temperature. As the imidazole group™s ionization state is a key determinant of intracellular
protein function, advocates of alpha-stat management contend that this strategy promotes
normal protein charge states and function, even at low temperatures.
The pH-stat approach increases the total carbon dioxide content of the blood as the tem-
perature falls in order to maintain fixed temperature-corrected pH values. The optimal pH of
most enzymatic reactions does vary with hypothermia, mostly in accordance with the predic-
tions of the alpha-stat hypothesis. Hence, the relative acidosis of pH-stat would be expected to
lower enzymatic reaction rates. Whether this is beneficial in reducing energy consumption,
or harmful by impairing key cellular homeostatic mechanisms, is unclear.
Differences in alpha-stat and pH-stat management become progressively greater as tem-
perature is reduced so the effect is quite profound below 25°C, but above 32°C, quantitatively,
the change in CO2 solubility is small and of much less clinical and physiological relevance.
This is further evident when one appreciates how little CPB time most adult cardiac surgical
patients spend at hypothermic temperatures. Most cases are conducted with mild hypother-
mia and in those much of CPB time is spent transitioning to, or from, those temperatures; the
actual time on CPB spent below 32°C may only be 25% of the total CPB time. Thus, although
frequently discussed, alpha-stat versus pH-stat management is of little actual relevance in
most adult cardiac surgery.
Chapter 6: Metabolic management

Potassium (K+)
Hyperkalemia is the most common electrolyte disturbance during CPB. Potassium levels can
be lowered using diuretics, insulin and dextrose administration, or hemofiltration. The treat-
ment of choice is dictated by the potassium level, the persistence of rise in potassium levels
and the presence or absence of electrophysiological disturbances. Serum potassium levels
transiently rise with the administration of cardioplegia and this will usually correct without
treatment within a short period after ceasing delivery of cardioplegia. Potassium levels in the
range 5.5“6.5 mmol/l can be treated with administration of a diuretic, usually furosemide
20“40 mg. In some centers, levels between 6.5 and 7 mmol/l are treated using insulin and dex-
trose infusions. Levels above 7 mmol/l or persistently raised potassium levels can be lowered
using “zero balance hemofiltration.” A crystalloid solution, typically normal saline, is added
to the CPB circuit to maintain circulatory volume and then removed by hemofiltration caus-
ing concomitant removal of potassium. As this technique can result in the loss of significant
amounts of bicarbonate through the hemofilter, it should be replaced using sodium bicar-
bonate titrated to blood bicarbonate levels.
The urgency or need to treat hyperkalemia should in part be determined by the presence
or absence of electrophysiological disturbance. In the absence of ECG changes, moderate
hyperkalemia may not require treatment. If treatment is chosen, its effect should not be longer
than the anticipated period of hyperkalemia. It is important to note that during CPB extracel-
lular potassium may rise but typically, even untreated, increases in K+ levels are nearly always
transient, as the extracellular potassium concentration in the plasma is quite small relative
to the intracellular capacity for its uptake. Rapid shifts to the intracellular space and urinary
excretion often correct K+ levels quite quickly after CPB.
Hypokalemia, usually less than 4.5 mmol/l, is treated by administration of potassium chlo-
ride, normally in 10“20 mmol boluses. It is worth bearing in mind that rapid bolus adminis-
tration of potassium during CPB may cause transient vasodilatation. Potassium levels alter
with temperature. Treatment should be undertaken in the context of:
• temperature;
• the rate of rise of the potassium level;
• the persistence of that level; and
• the point during surgery at which it is occurring.
Ideally, potassium is finally corrected before separation from CPB using results of electro-
lyte measurements taken at a body temperature of not less than 35°C.

Calcium (Ca2+)
Calcium levels are reduced by hemodilution, chelation by preservatives in bank blood or by
hemofiltration. Low serum Ca2+ levels are generally corrected close to the termination of CPB,
when the aortic cross-clamp has been removed, a cardiac rhythm has been established and the
temperature is approaching normothermia. One gram (or 3“5 mg/kg) of calcium chloride is
usually all that is required to normalize serum ionized calcium levels (1“1.5 mmol/l). Admin-
istration of Ca2+ may exacerbate reperfusion injury and should be avoided immediately before
or after cross-clamp removal. Timing of administration can be guided by normalization of
cardiac conduction indicating adequate reperfusion.
Chapter 6: Metabolic management

Magnesium (Mg+)
Magnesium depletion occurs during CPB if hemofiltration is used or if there is high volume
diuresis, particularly with loop diuretics. In these situations, a 2 mg bolus of Mg+ may be
added empirically into the circuit after the core temperature has reached 34°C and the aortic
cross-clamp has been removed. Ideally, if Mg+ levels are available, Mg+ administration should
be titrated according to blood levels.

Phosphate levels are commonly low after major cardiac surgery. This frequently occurs in
the immediate postoperative period and is associated with significant respiratory and car-
diac morbidity. Therefore, phosphate levels should be routinely measured after surgery, espe-
cially in patients with a complicated or prolonged intraoperative course, so that appropriate
replacement therapy may be started in a timely manner.

Phosphate levels on CPB tend to increase as a result of the physiological stress response to
major surgery. Values may exceed 20 mmol/l in diabetic patients without treatment. Non-dia-
betic patients™ serum glucose levels can also rise; levels of 10“15 mmol/l are not uncommon.
Continuous insulin infusions of 5“15 U/hour may be required during CPB. Hyperglycemia
is associated with poor patient outcomes. Specifically, perioperative hyperglycemia has been
associated with higher incidences of mediastinitis, wound infections and neurocognitive defi-
cits. Conflicting literature regarding both the ideal and acceptable intraoperative and postop-
erative glucose levels exists. However, recent studies have shown mixed results from attempts
at aggressive management of CPB-related hyperglycemia. The results range from favorable
outcomes, to little or no association between reducing serum glucose levels and reduction in
postoperative complications, to adverse patient outcomes associated with the tight control of
CPB-related hyperglycemia. It is generally believed that normal (4.0“5.5 mmol/l) serum glu-
cose levels during CPB are ultimately desirable. Consistent achievement of this goal remains
elusive at this time. Postoperative hypoglycemia, equally as dangerous and undesirable as
hyperglycemia, can result from clinicians “overshooting” in their attempts at serum glucose

Lactate is a major end product of glucose metabolism and gives an indication of the metabolic
status during CPB. Most patients exhibit a progressive increase in plasma lactate during CPB.
Lactate levels increase two- to threefold during normothermic and hypothermic CPB. Dur-
ing periods of hypoperfusion or decreased liver function, usually secondary to hypothermia,
serum lactate levels can increase even further (four- to eightfold). Re-warming the patient
and increasing flow rates usually helps to lower lactate levels.

Hemofiltration (Ultrafiltration)
This allows selective separation of plasma water and low-molecular-weight solutes from
the blood™s cellular and plasma protein. Hemofilter membranes are composed of thousands
of semipermeable hollow fibers (polysulfone, polyacrylonitrite or cellulose acetate fibers),
each with an internal diameter of ˜200 μm. Hollow fiber pore size determines which plasma
Chapter 6: Metabolic management

solutes will be removed. Pore size usually ranges between 10 and 35 angstroms, removing
molecules ¤20 000 Daltons. The sieving coefficient is a measure of hemofilter efficiency and is
directly related to solute molecular size. Solutes with weights ¤11 000 Daltons (Na+, K+, Ca2+,
Mg+, urea, creatinine, chloride, phosphorous, HCO3“, C3a, C5a, IL-1, IL-6, TNF-±) have a
sieving coefficient of 1, indicating they are filtered at the same concentration as they exist in
the blood. Larger molecules >20 000 Daltons (hemoglobin, globulin, fibrinogen, blood cells,
platelets, albumin and clotting factors) are unable to pass through the hemofilter fiber pores.
The hydrostatic pressure differential, or transmembrane pressure (TMP), across the hemo-
filter rather than the osmotic pressures, as in hemodialysis, creates the separation of solutes
and fluids. Application of vacuum to the effluent side of the hemofilter will improve solute
and fluid filtration rates (up to 180 ml/minute). TMP (100“500 mmHg) is the mean of the
hemofilter inlet (PI) and outlet (PO) pressures plus vacuum (PV): TMP=PI + PO/2 + PV. The
combination of these pressures determines the filtration rate.

Drug dilution and loss
The CPB circuit adds as much as 1800 ml to the adult patient™s circulating volume, especially
in institutions that do not employ prime-reducing techniques (AP, vacuum-assisted venous
drainage and microcircuitry).
Institution of CPB results in an immediate dilution of drug concentrations. A new equi-
librium between protein-bound and free ionized drug concentrations is established. Drug
clearance and the intensity of biological effect are proportional to the concentration of free
(unbound) drug, thus the pharmacodynamic effects of drugs are not necessarily altered if the
concentration of free drug is maintained.
The effect of CPB on drug concentrations is complex and also influenced by a number of
factors such as temperature and the type of materials used in the CPB circuit; certain types
of plastics and coatings on oxygenator membranes are more prone to binding drugs than

Monitoring of patient metabolic and physiological parameters
Arterial and venous blood gases and electrolytes
In-line real-time blood gas analysis has become the “gold standard” for extracorporeal per-
fusion. These arterial and venous analyzers utilize disposable sensors that attach directly to
the arterial and venous lines of the CPB circuit, providing a luminal surface interface for
blood leaving the oxygenator and for blood returning from the patient. Arterial sensors pro-
vide pH, PO2, PCO2, BE, HCO3“ and SaO2 data. The venous sensor generally provides HCT,
hemoglobin and mixed venous oxygen saturation (SvO2) measurements. Some in-line devices
will also provide a continuous calculation of oxygen consumption based on pump flow and
the arterio-venous oxygen differential.
If in-line blood gas monitoring is not available, intermittent samples should be taken at
30-minute intervals for analysis. Most blood gas machines provide data on blood gases, acid“
base status, hematocrit, hemoglobin, electrolytes and glucose.

During CPB SvO2 is an indicator of the matching of DO2 and VO2. As the margin between
systemic O2 delivery and demand narrows, O2 extraction increases and SvO2 is reduced.
Chapter 6: Metabolic management

Reduced depth of anesthesia or degree of muscular paralysis by muscle relaxant drugs,
low inspired oxygen concentration in the fresh gas flow mixture or anemia all decrease SvO2.
However, if these parameters have been optimized, low SvO2 values generally indicate hypo-
perfusion and should prompt an increase in pump flow rate to improve oxygen delivery. If the
ability to increase flow is limited by venous return, then increasing DO2 by increasing HCT, or
reducing VO2 by reducing temperature, is indicated. However, the SvO2 value should always
also be interpreted in the context of core temperature. The solubility and hemoglobin bind-
ing affinity of oxygen increases with hypothermia, whilst organ metabolic demand decreases,
resulting in increased SvO2 if perfusion is adequate. Venous saturations of 65“75% are typical
at temperatures of 37“35°C, 76“85% at temperatures of 34“32°C, and 85“100% at tempera-
tures of 32“16°C.

• Grocott HP, Mackensen GB, Grigore AM,
Suggested Further Reading et al. Postoperative hyperthermia is
• Butterworth J, Wagenknecht LE, Legault C, associated with cognitive dysfunction after
et al. Attempted control of hyperglycemia coronary artery bypass graft surgery. Stroke
during cardiopulmonary bypass fails to 2002; 33: 537“41.
improve neurologic or neurobehavioral
• Mackensen GB, Grocott HP, Newman MF.
outcomes in patients without diabetes
Cardiopulmonary bypass and the brain. In
mellitus undergoing coronary artery bypass
Kay PH, Munsch CM, eds. Techniques in
grafting. J Thorac Cardiovasc Surg 2005;
Extracorporeal Circulation, 4th ed. London:
130: 1319.
Oxford University Press; 2004: 148“76.
• Gandhi GY, Nuttall GA, Abel MD, et al.
• McAlister FA, Man J, Bistritz L, et al.
Intensive intraoperative insulin
Diabetes and coronary artery bypass
therapy versus conventional glucose
surgery: an examination of perioperative
management during cardiac surgery: a
glycemic control and outcomes. Diabetes
randomized trial. Ann Intern Med 2007;
Care 2003; 26: 1518“24.
146: 233“43.
• Puskas F, Grocott H, White W, et al.
• Gravlee GP, Davis RF, Stammers AH,
Intraoperative hyperglycemia and cognitive
Ungerleider R, eds. Cardiopulmonary
decline after CABG. Ann Thorac Surg 2007;
Bypass: Principles and Practice, 3rd ed.
84: 1467“73.
Philadelphia: Wolters Kluwer Health/
• Reed CC, Stafford TB. Cardiopulmonary
Lippincott Williams & Wilkins; 2008.
Bypass, 2nd ed. Houston: Texas Medical
• Grigore AM, Grocott HP, Mathew JP, et al.
Press; 1985.
The rewarming rate and increased peak
• Watkins, JG. Arterial Blood Gases:
temperature alter neurocognitive outcome
A Self-Study Manual. Philadelphia:
after cardiac surgery. Anesth Analg 2002; 94:
Lippincott Williams and Wilkins; 1985.

Myocardial protection and

7 Constantine Athanasuleas and Gerald D. Buckberg

Optimal outcomes after cardiac surgery are dependent on several factors; primarily these
are appropriate patient selection, precise surgical technique and intraoperative protection
of viable myocardium. Myocardial damage may be associated with cardiopulmonary bypass
(CPB) to varying degrees and influences morbidity and mortality. Cardioplegia, the solution
used to protect the myocardium intraoperatively, in the period during which the heart is
isolated from the circulation and hence ischemic, may differ in composition, route of admin-
istration and method of delivery to the heart. This chapter reviews the rationale for the use of
cardioplegia, techniques of administration, components of different cardioplegia solutions
and applications of cardioplegia in different surgical interventions.

Myocardial damage during cardiopulmonary bypass
Damage to the myocardium can occur during short or long operations, but is more likely
under certain clinical circumstances. These include prolonged aortic cross-clamp times,
impaired ventricular function, concomitant valve or aortic surgery with coronary bypass,
re-operation and operation during acute coronary ischemia.
During CPB, the aorta is usually clamped to provide a dry operative field with good vis-
ibility. If the heart continues to beat during aortic cross-clamping, intracellular depletion
of high-energy phosphates ensues and results in impaired recovery of function. Intermit-
tent cross-clamping with multiple periods of reperfusion has been shown to be suboptimal
because episodes of reperfusion injury may occur following each release of the aortic clamp.
The technique of cross-clamp and fibrillation, as an alternative method to using cardiople-
gia to provide a bloodless field and non-beating heart, engenders multiple short periods of
ischemia followed by reperfusion. This may further potentiate, rather than prevent, ischemic
damage, as redistribution of flow away from the vulnerable subendocardial muscle occurs.
Postoperatively, myocardial damage can be detected by electrocardiography, echocar-
diography, radioactive imaging studies and cardiac magnetic resonance imaging. Chemical
markers of damage include troponin or creatine phosphokinase release. The clinical manifes-
tations of myocardial damage may present as low cardiac output syndrome due to impaired
myocardial contractility, dysrhythmias, decreased ventricular compliance or segmental myo-
cardial wall motion abnormalities.

Goals and principles of myocardial protection
The goal of myocardial protection with cardioplegia is to prevent myocardial injury during
the periods of intentional ischemia that are required to perform cardiac operations. This can
be accomplished by adjusting myocardial metabolic requirements both during the phase of
no perfusion, and following reperfusion, in such a way as to minimize the deleterious effects
of prolonged ischemia.
80 Cardiopulmonary Bypass, ed. S. Ghosh, F. Falter and D. J. Cook. Published by Cambridge University Press.
© Cambridge University Press 2009.
Chapter 7: Myocardial protection and cardioplegia

Figure 7.1 (a) Left ventricular oxygen requirements of the beating, empty, fibrillating, arrested heart from 37°C
to 22°C. Note the lowest requirements during arrest. (b) The left ventricular oxygen requirements of a beating/
working heart and an arrested heart at 37°C, 22°C and 10°C.

The principal determinants of myocardial energy utilization are left ventricular end-
diastolic wall tension (LVEDP) and electromechanical activity; limitation of both of these
parameters can thus limit myocardial metabolic demand.
During diastole, with low LVEDP, myocardial oxygen consumption and energy substrate
utilization is minimized. The rapid attainment of diastolic cardiac arrest at the onset of the
ischemic period following application of the aortic cross-clamp effectively places the heart in
a state of hibernation, particularly if the myocardium is simultaneously cooled.
Electromechanical activity raises oxygen demand during ischemia. Therefore, ideally, all
electrical activity must cease during cardioplegia-induced cardiac arrest. Hypothermia, which
in itself lowers basal metabolic rate and thus helps to reduce myocardial electrical activity, was
used in thousands of operations in the early years of cardiac surgery. Topical cold saline or
iced slush was poured over the heart into the pericardial cavity after aortic cross-clamping.
Though effective for shorter periods of cross-clamping, this method becomes limiting follow-
ing ischemic periods exceeding 1 hour. Subsequently, infusion of cold cardioplegia solutions
into the myocardium to rapidly stop electromechanical activity and simultaneously reduce
temperature in all myocardial layers has become the preferred method of cooling the heart
(see Figure 7.1).
Interventions that maximize high-energy phosphate production, while minimizing high-
energy phosphate utilization and intracellular calcium accumulation during ischemia and
reperfusion, are effective in delaying or preventing the onset of ischemic necrosis and in
aiding recovery of function following reperfusion. Examples of methods to maximize high-
energy phosphate production include preoperative glucose and glycogen loading, intraopera-
tive infusions of glucose, insulin and potassium, or the addition of Kreb™s cycle intermediates,
glutamate and aspartate, to cardioplegic solutions.

Components of cardioplegia
The constitution of cardioplegia solutions varies according to individual surgeons and insti-
tutional preferences. The composition of cardioplegic solutions has been described as being
similar to either the ionic composition of extracellular or of intracellular fluid depending on
the content of sodium, potassium, calcium and magnesium of a given type of cardioplegia.
Cardioplegia solutions can be further categorized according to whether they are crystalloid or
Chapter 7: Myocardial protection and cardioplegia

blood based. The essential requirement for attainment of rapid diastolic cardiac arrest, how-
ever, renders potassium (20“40 mEq/l), which causes membrane depolarization, an essential
ingredient of all cardioplegia solutions.
Other components common to cardioplegia solutions include sodium (100“200 mEq/l)
and chloride ions. Sodium minimizes the transcellular sodium gradient and so reduces
intracellular edema; marked extracellular hyponatremia (<50 mEql), together with excessive
potassium-induced membrane depolarisation, alters the Na+/Ca2+ exchange mechanisms in
such a way as to promote intracellular Ca2+ accumulation causing damage to sarcolemma
membranes. Chloride ions maintain the electroneutrality of the solution.
Modification of cardioplegia to produce a solution that provides optimal preservation of
myocardial function has led to a variety of additions to the “basic” ingredients, for example
one of the most established blood cardioplegia preparations contains citrate phosphate dex-
trose (CPD), to limit calcium influx during ischemia, and tromethamine (tris-hydroxymethyl
aminomethane, THAM), a buffer that prevents acidosis. THAM diffuses into the intravas-
cular space, “captures” the CO2 produced by metabolic acidosis and improves myocardial
The original cardioplegic solutions used for many years consisted of crystalloid solutions
with various additives. The most widely used is the St. Thomas™ Hospital solution. Calcium, in
low concentration, is included in the solution to ensure that there is no likelihood of calcium
paradox during reperfusion and to maintain integrity of cell membranes. Magnesium may
help stabilize the myocardial membrane by inhibiting a myosin phosphorylase, which protects
ATP reserves for postischemic activity. The protective effects of magnesium and potassium
have been shown to be additive. Procaine, a local anesthetic, is included in low concentration,
to counteract the vasocontrictive effects of particulate contaminants in the infusion and so
promote even distribution.
St. Thomas™ solution is usually buffered by the addition of sodium bicarbonate just prior
to use; this renders the solution slightly alkaline and helps compensate for the metabolic aci-
dosis that accompanies ischemia. Commercially prepared bags of ready diluted St. Thomas™
cardioplegia are available as an alternative to diluting the concentrate with Ringer™s, but differ
slightly from the original St Thomas™ preparation.
Hypothermic crystalloid cardioplegia has certain disadvantages, including the fact that
it inhibits the enzyme Na+/K+ adenosine triphosphatase, which is intrinsic to the function of
transmembrane ion pumps, thereby producing myocardial edema and consequent activation
of platelets, leukocytes and complement.
Blood cardioplegia largely replaced crystalloid cardioplegia in most centers several years
ago. It consists of four parts of blood to one part crystalloid cardioplegia solution. This lim-
its the systemic hemodilution seen with crystalloid cardioplegia during repeated infusions.
Blood cardioplegia maintains oncotic pressure, is a natural buffering agent, has advantageous
rheological properties and is a free radical scavenger. It also limits reperfusion injury in the
acutely ischemic myocardium. Experimental studies have shown that normal hearts sub-
jected to up to 4 hours of ischemia have complete recovery of function when intermittent cold
blood cardioplegia is infused. Cold blood cardioplegia alone, however, does not totally avoid
injury. The Kreb™s cycle amino acids glutamate and aspartate are depleted during episodes of
intermittent blood cardioplegia administration. They are especially depleted in chronically
ischemic hearts and may be replenished by using blood cardioplegia with added glutamate
and aspartate, often referred to as “substrate-enhanced cardioplegia.” Blood cardioplegia,
with or without substrate enhancement, may be infused as a warm solution to optimize the
Chapter 7: Myocardial protection and cardioplegia

1.5 1.5
Control 37°C
Cold blood
(4 hours)
1.0 1.0

Normothermic glutamate
(gm-m/kg) (gm-m/kg)
(45 min)
0.5 0.5 4°C

5 10 15 20 25
10 15 20 25
(a) (b)
LAP (mmHg) LAP (mmHg)
Figure 7.2 (a) Left ventricular function in normal hearts subjected to 4 hours of aortic clamping with blood
cardioplegia every 20 minutes compared with depressed function after 45 minutes of normothermic arrest without
cardioplegia. (b) Left ventricular function when jeopardized hearts undergoing 45 minutes of normothermic
ischemia are subjected to 2 more hours of aortic clamping. Note (1) no further improvement when only cold
cardioplegic perfusate is given over the 45 minute arrest period, (2) progressively increased recovery when the
cardioplegic solution is supplemented with warm glutamate and aspartate during induction of cardioplegia and
reperfusion with intermittent cold doses of blood every 20 minutes of supplemental aortic clamping. LAP = left
atrial pressure; SWI = stroke work index.

Table 7.1. Composition of crystalloid (STH1) and blood-based St. Thomas™ Hospital cardioplegic (BSTH1)
solutions (concentrations delivered to heart)

Ionic composition STH1 BSTH1
Na+(mmol/l) 144 142
K+ (mmol/l) 20 20
Mg2+ (mmol/l) 16 16
Ca2+ (mmol/l) 2.2 1.7
HCO3“ (mmol/l) 0* 30“40
Procaine (mmol/l) 1 1
pH 5.5“7.0 7.4
Hematocrit 0 10“12%
Osmolarity (mOsmol/kg H2O) 300“320 310“330
Note: * NaHCO3 added prior to use, increasing HCO3“ and pH.
Note: These are high K+ solutions; low K+ solutions (10 mmol/l) may be used for additional doses.

metabolic rate of repair, just prior to removal of the aortic cross-clamp, at the end of the
intended ischemic period. This warm phase has been referred to as the “hot shot” of cardio-
plegia. It enhances cellular assimilation of the substrates (see Figure 7.2), augmenting the rate
of recovery of myocardial contractility. Some surgeons also infuse a small volume of warm
blood cardioplegia followed by cold cardioplegia to induce cardiac arrest at the commence-
ment of the ischemic period, on the basis that this “feeds” the heart, i.e., provides a more
physiological delivery of oxygen and substrates for the period of ischemia.
Chapter 7: Myocardial protection and cardioplegia

Table 7.2. Substrate-enhanced blood cardioplegia solution (high K+)

Additive Concentration delivered*
K+ (2 mEq/ml) 16“20 mmol/l
THAM (0.3 mol/l) pH 7.5“7.7
Citrate“phosphate“dextrose 0.2“0.4 mmol/l
Aspartate 13 mmol/l
Glutamate 13 mmol/l
Dextrose 50% <400 mg/l
Dextrose 5% 380“400 mOsm
* When mixed with blood in a 4:1 ratio.

Table 7.3. “Multi-dose” low-potassium cold blood cardioplegia solution

Additive Concentration delivered*
K+ (2 mEq/ml) 8“10 mmol/l
THAM (0.3 mol/l) pH 7.6“7.8
Citrate“phosphate“dextrose 0.5“0.6 mmol/l
Dextrose 5% 380“400 mOsm
* When mixed with blood in a 4:1 ratio.

Compositions of typical crystalloid and blood cardioplegia solutions are shown in Tables
7.1, 7.2 and 7.3.

Cardioplegia delivery
Cardioplegia delivery systems generally comprise an infusion system with in-line pressure
monitors, a cardioplegic heat exchanger for cold and warm perfusion, and cannulae for ante-
grade and retrograde delivery. For further details about cannulae and cardioplegia delivery
please see Chapter 1.

Routes of cardioplegia delivery
To be effective the desired volume of cardioplegic solutions must be evenly distributed
throughout the myocardium. Obstacles to this goal include coronary stenoses, which pre-
vent uniform delivery of cardioplegia, often to the most vulnerable regions of the myocar-
dium, and aortic regurgitation. Aortic regurgitation, even when mild, lowers aortic root
pressure, so reducing the perfusion pressure of cardioplegia infused into the aortic root,
and also causes loss of cardioplegia into the left ventricle. Hence, the time honored method
of antegrade aortic root infusion of cardioplegia alone can be combined with “retrograde
cardioplegia” infusion into the coronary sinus to overcome the potential for inadequate
myocardial preservation.
Retrograde coronary venous cardioplegia perfusion via the coronary sinus overcomes the
limitations of antegrade administration via the coronary arteries alone to ensure adequate left


. 3
( 8)