Perfusion Pressure during Cardiopulmonary Bypass

Selection of perfusion pressure during CPB is based on balancing the demands of surgical access (bloodless field) with patient outcome (adequate oxygen delivery). Lower flow and pressure during CPB may optimize visualization, whereas higher flow and pressure may minimize patient complications. Determining the optimum perfusion pressure has been extremely challenging because no single study can adequately address all the complexities of CPB. Because of the brain’s poor tolerance of ischemia, neurologic outcome has been the most common outcome studied in relation to perfusion pressure. The complicated relationship between neurologic outcome and perfusion pressure is likely related to two causes of adverse neurologic outcomes: hypoperfusion and embolism.

Between mean arterial pressures (MAP) of 50 and 150 mmHg, cerebral autoregulation maintains a relatively constant blood flow and oxygen delivery. During hypothermic CPB, the lower limit of cerebral autoregulation may be as low as 20 to 30 mmHg,² affording some additional protection against hypoperfusion. Increasing perfusion pressure to alleviate the risk of hypoperfusion may lead to greater embolic load and worse outcomes. Ultimately, the selection of perfusion pressure during CPB will need to be based on clinical outcome studies.

Subgroups at increased risk for adverse outcomes that may benefit from higher perfusion pressure during CPB include patients with severe atheromatous disease (cerebrovascular or aortic arch), advanced age, systemic hypertension, and diabetes. Increased cerebral dysfunction in the elderly may be a result of slower vasodilatation of cerebral resistance vessels during periods of rewarming and subsequent transient episodes of metabolism-flow mismatch with resultant ischemia. It is unknown what the effect of elevating perfusion pressure during rewarming would be on neurologic outcome. Hypertensive patients are generally accepted to have intact pressure-flow autoregulation, with a rightward shift in the cerebral autoregulation curve such that pressure-dependent flow patterns develop at higher perfusion pressures than in the normal population. In hypertensive patients the use of higher perfusion pressure during CPB is common practice. Patients with type 1 diabetes mellitus appear to have impaired metabolism-flow coupling during CPB. They also have some loss of pressure-flow autoregulation.

Once the CPB team has selected target perfusion pressures during CPB, a few technical issues emerge. Throughout this discussion perfusion pressure and MAP have been used almost interchangeably. In general, cerebral perfusion pressure is what is of most concern. Cerebral perfusion pressure is determined by the difference between MAP and the higher of central venous pressure and intracranial pressure. The latter values are usually less than 5 mmHg during CPB. However, in the presence of compromised cerebral venous drainage (malpositioned cannula, patient positioning), MAP will not accurately reflect cerebral perfusion pressure.

Measurement artifacts also play a role in perfusion pressure management. MAP may vary by as much as 20 mmHg over 30 seconds while pump flow is constant. The mechanism of this oscillation and its relation to outcome are unclear. A more common artifact is discordance between radial arterial and central arterial pressures during rewarming. This difference may be as great as 30% and is believed to occur from opening of arteriovenous shunts in the arm.

After acknowledging the technical issues of pressure monitoring, the CPB team is left to maintain the selected perfusion pressure. To achieve these perfusion pressure goals the team has two general options: alterations of pump flow or administration of vasoactive agents. Increasing pump flow may be used as a temporizing measure for hypotension if surgical demands allow it; however, this may come at the cost of dangerously reducing reservoir volume. Alternatively, phenylephrine and norepinephrine may be used to support perfusion pressure. In the case of hypertension, pump flow may be reduced, although this increases the potential for inadequate oxygen delivery; more commonly, a vasodilator, such as sodium nitroprusside or nitroglycerin, is employed. Isoflurane or another volatile anesthetic may be administered through the pump oxygenator, with careful attention paid to its use during weaning from CPB.

Nonpulsatile versus Pulsatile Perfusion

It remains uncertain whether pulsatile CPB improves outcome compared with standard, nonpulsatile CPB. Claims of advantages to pulsatile flow are effectively offset by conflicting studies of similar design.

The comparatively small size of the arterial inflow cannula can effectively filter out a large component of the pulsatile kinetic energy. Consequently, as achieved clinically, pulsatile flow may actually be quite similar energetically to nonpulsatile flow. This potentially unrecognized lack of difference in types of flow may partly explain the failure of pulsatile perfusion to alter hormone levels and clinical outcome.

Pump Flow during Bypass

Like perfusion pressure, pump flow during CPB represents a careful balance between the conflicting demands of surgical visualization and adequate oxygen delivery. Two theoretical approaches exist. The first is to maintain oxygen delivery during bypass at normal levels for a given core temperature. Although this may limit hypoperfusion, it does increase the delivered embolic load. The second approach is to use the lowest flows that do not result in end-organ injury. This approach offers the potential advantage of less embolic delivery as well as potential improved myocardial protection and surgical visualization.

During CPB, pump flow and pressure are related through overall arterial impedance, a product of hemodilution, temperature, and arterial cross-sectional area. This is important because the first two factors, hemodilution and temperature, are criticaldeterminants of pump flow requirements. Pump flows of 1.2 L/min/m² perfuse most of the microcirculation when the hematocrit is near 22% and hypothermic CPB is being employed. However, at lower hematocrits or periods of higher oxygen consumption these flows become inadequate.

Most perfusion teams also monitor mixed venous saturation, targeting levels of 70% or greater. Unfortunately, this level does not guarantee adequate perfusion of all tissue beds, because some (muscle, subcutaneous fat) may be functionally removed from circulationduring CPB. Hypothermic venous saturation may overestimate end-organreserves. Regionalperfusion of various end-organs (brain, kidney, small intestine, pancreas, and muscle) has been quantified with a fluorescent microsphere technique.³ Cerebral blood flow was unchanged at higher pump flows. Renal perfusion was maintained at flows of 1.9 and 1.6 L/min/m². Perfusion to the pancreas was constant at all flows, and small bowel perfusion varied linearly with pump flow. Muscle bed flows were decreased at all flows.

During CPB, most of the outcomes studied in relation to pump flow are those related to the organs at high risk for ischemic injury (i.e., kidney and brain). Much work has been applied to examining the relationship between renal dysfunction and pump flow. Preexisting renal disease is a consistent predictor of postoperative renal dysfunction, the incidence of which ranges between 3% and 5%. Renal function appears unaltered when pump flows greater than 1.6 L/min/m² are employed, but whether this management will affect outcomes in patients with preexisting renal dysfunction is less clear.

Bypass Temperature Management Strategy

Although hypothermic temperatures have been employed since the advent of extracorporeal circulation, the importance of reduced temperatures during bypass was challenged in the early 1990s.

Temperature Monitoring

Because the brain is vulnerable to hyperthermic temperatures, it is important to use the temperature-monitoring site most likely to reflect cerebral temperature. The most commonly used sites in cardiac surgery patients include esophageal, nasopharyngeal, tympanic, pulmonary arterial, rectal, urinary bladder, subcutaneous (or muscle), and cutaneous sites. Unfortunately, none of these monitoring locations has been demonstrated to reflect cerebral temperature reliably. With exposure of the brain, investigators have placed a thermocouple directly in the cerebral cortex. Brain temperature was compared with values obtained from sensors in eight locations.⁵ Investigators found a poor concordance between cerebral temperature and values obtained at the other monitoring sites. Locations hypothesized to best reflect core temperature—tympanic membrane, esophagus, nasopharynx, pulmonary artery—sometimes overestimated cerebral temperature or underestimated brain temperature. Because of the substantial variability noted in central temperature readings (Fig. 22-2) and lack of the concordance of central temperature measures in every patient, the investigators recommended the use of at least three measures of central or core temperature.

Figure 22-2 Graphic depiction of temperature-time relationship during cardiopulmonary bypass and deep hypothermic circulatory arrest. Temperatures are plotted every minute. Central sites are the nasopharynx, tympanic membrane, esophagus, and pulmonary artery. Peripheral sites are the bladder, rectum, axilla, and sole of the foot. The two slowest cooling central sites were tympanic membrane and esophagus.

(From Stone JG, Young WL, Smith CR, et al: Do standard monitoring sites reflect true brain temperature when profound hypothermia is rapidly induced and reversed? Anesthesiology 82:344, 1995.)

Acid-Base Strategy

The management of acid-base status during hypothermic CPB has been a long-standing source of debate. Understanding of the physiologic responses to hypothermia, and the influences of PCO₂ have led to shifts in clinical practice over the past decades. Two strategies exist for managing acid-base balance during periods of hypothermia: α-stat and pH-stat. The term α-stat was first proposed to describe the theory that acid-base regulation in vertebrate animals functioned during temperature fluctuation to maintain a constant ratio (α) of dissociated to undissociated forms of the imidazole ring on histidine. It is this protein charge state that is important in regulating pH-dependent cellular processes. Hypothermia increases the solubility of oxygen and carbon dioxide in the blood, leading to a decrease in PCO₂ and an increase in pH at lower temperatures. With α-stat blood gas management, the uncorrected (37°) pH is kept at 7.40 with the PCO₂ at 40 mmHg, creating a relative alkalosis at the patient’s actual body temperature. This strategy is considered to be physiologic because the ionization state of histidine is unchanged over all temperature ranges and protein structure and function are preserved.

The pH-stat approach to acid-base balance maintains a pH of 7.40 and PCO₂ of 40 mmHg when corrected for body temperature, typically requiring the addition of CO₂ during hypothermic CPB. This method of blood gas management was generally favored until the mid 1980s because it was believed that the potent vasodilatory effects of CO₂ would provide increased cerebral blood flow and thereby minimize the risk of cerebral ischemia during CPB. It is now recognized that pH-stat management during hypothermia produces passive cerebral vasodilation, impairs autoregulatory responses to blood pressure changes and metabolic demands in the brain, and does not improve overall oxygen balance. In contrast, α-stat management preserves autoregulation and the relationship between cerebral blood flow and metabolism. Neither blood gas strategy has any significant effect on hypothermic cerebral metabolism. The increased CBF seen with pH-stat may also increase the risk of cerebral embolization or produce a steal phenomenon.⁶

Fluid Management

The choice of fluid for priming the extracorporeal circuit in CPB remains controversial. The idea of using nonblood prime was first introduced in 1959. This technique of hemodilution was found to be safe when combined with hypothermia to reduce oxygen consumption and demand. The use of nonblood primes and moderate hemodilution for CPB has become routine in most centers. A reduction in hematocrit from 40% to 20% allows cooling to 22°C without an increase in blood viscosity or required driving pressure. Hematocrit reduction may be achieved before bypass by means of acute normovolemic hemodilution in the hope of reinfusing the patient’s own heparin-free blood, rather than allogeneic red blood cells, after CPB.

Several studies have investigated the differences between colloid and crystalloid priming solutions. In general, crystalloid solutions lead to decreased colloid osmotic pressure with a resultant increase in extracellular water retention, irrespective of the osmolarity of the pump prime. Albumin, unlike a pure crystalloid prime, can decrease the interaction of blood components with the bypass circuit by coating the fluid pathway surfaces. In their meta-analysis of 21 controlled trials enrolling 1346 patients, Russell and associates showed a notably smaller drop in on-bypass platelet counts in patients treated with albumin in the pump prime.⁷

Ultrafiltration during bypass can be used as a means of reducing excess water accumulation. Modified ultrafiltration describes the process of hemofiltration immediately after the cessation of bypass. This process results in a more consistent reduction in total body water with significant increases in hematocrit, myocardial contractility, cardiac index, and improved pulmonary compliance.

Myocardial Injury

Most coronary revascularization procedures are completed with the assistance of CPB. Although the completion of coronary anastomoses is facilitated by CPB (i.e., the surgeon can operate on a quiet, nonbeating heart), the heart is subjected to a series of events leading to ischemic myocardium during extracorporeal circulation. The operation, which is designed to preserve and improve myocardial function, is sometimes associated with myocardial damage (Box 22-2). The extent and incidence of this injury are dependent on the sensitivity and specificity of the diagnostic methods being used. However, most patients who undergo cardiac operations sustain some degree of myocardial injury. Although patients with normal ventricular function may tolerate these minor amounts of injury without detectable sequelae, those with impaired ventricular function preoperatively may not be able to tolerate the slightest injury. As the patient population for CPB continues to become older and have greater degrees of concomitant illness, understanding the physiology of and developing effective preventive strategies for myocardial injury during CPB are increasingly important. Because myocardial damage influences early and long-term results, the identification and control of factors associated with myocardial injury are critical to ensuring good outcomes. Although injury may be linked to anesthetic and surgical management, myocardial injury usually is thought to occur from inadequate myocardial protection during CPB.

Brain Injury

The brain is highly susceptible to injury during CPB. Many clinicians believe that cerebral injuries after cardiac surgery are the most devastating adverse outcomes associated with CPB. A study of 2400 patients undergoing elective coronary artery bypass grafting (CABG) from 24 U.S. centers reported that 6.1% of patients suffer adverse postoperative gross neurologic or psychiatric central nervous system events. These patients remain in the intensive care unit and hospital for greater periods of time, and 1 in every 3 surviving patients does not return home but requires continued long-term care and rehabilitation.⁸ (see chapter 23).

Renal Dysfunction

The effects of CPB on the renal system have significant health and economic impacts; however, despite intensive investigation into the pathogenesis and prevention of renal failure, there remains limited progress in the development of effective protective strategies in recent decades.⁹ Because intravascular volume depletion and hypoperfusion can lead to exacerbation of renal ischemia and accentuate the risk for postoperative acute renal failure, avoidance of nephrotoxic agents and close attention to intravascular volume, blood pressure, and cardiac output (CO) are central in the effort to reduce the occurrence of acute renal failure after cardiac surgery.

Gastrointestinal Effects

The effects of CPB on the gastrointestinal system are complex and not fully elucidated. Although most patients undergoing cardiac surgery do not suffer adverse changes in gastrointestinal function, subclinical perturbations including transient elevations in hepatocellular enzymes and hyperamylasemia have been observed after CPB. Although the incidence of gastrointestinal complications after CPB is low (range, 0.3% to 3.7%), they are associated with significant morbidity and remarkably high mortality (range, 11% to 67%) compared with cardiac surgery patients without postoperative gastrointestinal compromise. The frequently reported adverse gastrointestinal outcomes include gastroesophagitis, upper and lower gastrointestinal hemorrhage, hyperbilirubinemia, hepatic and splenic ischemia, colitis, pancreatitis, cholecystitis, diverticulitis, mesenteric ischemia, as well as intestinal obstruction, infarction, and perforation.¹⁰

Although the pathophysiology of gastrointestinal complications after cardiac surgery is likely multifactorial, a unifying mechanism is splanchnic hypoperfusion. The gastrointestinal system is particularly vulnerable for ischemia due to the lack of autoregulation and to the preferential shunting of blood away from the gastrointestinal circulation during periods of hypotension. Hypothermia and nonpulsatile flow during CPB may be detrimental to mucosal perfusion. However, hypothermia has little effect on hepatic arterial blood flow and may actually increase portal flow. There is no significant difference in hepatic blood flow between pulsatile and nonpulsatile perfusion at high flow rates (2.4 L/min/m²) during hypothermia. Perhaps more important to the development of inadequate gastrointestinal perfusion is the significant increase in total body oxygen consumption in the immediate hours after CPB. Visceral hypotension is the most significant factor in the development of gastrointestinal complications after cardiac surgery. Gut ischemia of sufficient duration impairs gastrointestinal tract barrier function. Studies evaluating gut permeability have shown that CPB is associated with an increase in mucosal permeability and systemic endotoxin concentration.

Endocrine and Inflammatory Responses