Vasopressin, or antidiuretic hormone (ADH), secreted by the posterior pituitary gland, is a potent regulator of renal water excretion (17). At high concentrations, ADH may increase peripheral vascular resistance and decrease cardiac contractility and coronary blood flow (17,18). Animal studies have shown that infused arginine vasopressin (AVP) decreases cardiac oxytocin receptor expression and increases diastolic dysfunction in induced ischemia animal models (19,20). However, post-cardiac-surgery patients showed no cardiac dysfunction with AVP infusions (21,22). Similarly, in septic patients, those who received AVP required less vasopressors compared to controls while having no increase in mortality (23). ADH increases renal vascular resistance, reducing renal blood flow. ADH stimulates the release of the von Willebrand factor, perhaps improving hemostasis during and after cardiac surgery (see Chapter 22). Stimuli provoking ADH release include increased plasma osmolality, decreased blood volume or blood pressure, hypoglycemia, angiotensin, stress, and pain (17). General anesthesia and surgery are associated with moderate increases in ADH (24,25), and angiotensin-converting enzyme (ACE) inhibitors, which are commonly administered to cardiac surgical patients, have been associated with the syndrome of inappropriate antidiuretic hormone (SIADH) secretion (26). Cardiac surgery with CPB is associated with striking increases in ADH concentration, far above those seen during other major surgical procedures, and these effects may persist for hours postoperatively (25,27,28,29,30) (Fig. 11.2).

The exaggerated ADH response to CPB could be initiated by any number of stimuli, including the decrease in circulating blood volume upon initiating bypass. Left atrial pressure decreases markedly, especially with left-ventricular venting, thereby simulating volume depletion, which is a potent stimulus for ADH release. The transient hypotension normally occurring at the onset of bypass could stimulate increased ADH secretion. Pulsatile perfusion during CPB attenuates the exaggerated ADH response, particularly after bypass, but does not eliminate it (28,30,31) (Fig. 11.3). Pulsatile perfusion does not seem to significantly increase urinary output, despite reduced ADH concentrations (30).

Preoperatively, ADH levels were increased in patients with low left-ventricular ejection fraction (EF) or higher New York Heart Association (NYHA) heart failure class, and higher preoperative ADH concentrations served paradoxically as a predictor for postoperative vasoplegia. Plasma ADH concentrations did not increase nearly as much postoperatively in the vasoplegic group (32,33). This finding is consistent with the concept that vasoplegic syndrome is related to inappropriately reduced ADH concentrations in those patients that had increased concentrations preoperatively, which may reduce the ability to increase ADH as part of the stress associated with cardiac surgery (32). A reduced EF and/or chronic treatment with ACE inhibitors were independently associated with vasoplegia and with this relative postoperative ADH deficiency (33). This finding is consistent with studies in sepsis associating inappropriately diminished ADH levels in the presence of chronic stimuli to ADH secretion (34).

Certain anesthetic techniques, for example, maintenance of anesthesia with large doses of synthetic opioids (fentanyl or sufentanil) or with regional anesthesia, attenuate the hormonal responses associated with surgical procedures. Indeed, Kuitunen et al. (35) found that patients anesthetized with 50 µg/kg fentanyl demonstrated significantly lower AVP concentrations after CPB than patients who received a lighter plane of general anesthesia using inhaled enflurane. However, even opioid anesthesia will not completely ablate ADH release at the onset of CPB (29). Unfortunately, multiple studies provide conflicting data as to whether higher peak ADH concentrations occur during and after CPB in patients undergoing coronary artery surgery or valve surgery (27,28,30) (Fig. 11.2). In summary, ADH concentrations increase markedly during CPB irrespective of the anesthesia or perfusion technique, and the level of increase may associate with the likelihood that a patient will develop vasoplegia.

The catecholamines epinephrine and norepinephrine are products of the adrenal medulla and (in the latter case) of peripheral sympathetic and central nerve terminals. Marked elevations of plasma epinephrine and norepinephrine concentrations occurring during CPB underlie many hemodynamic sequelae of bypass, including peripheral vasoconstriction and shifts in intraorgan blood flow (31,36,37,38,39,40,41). With hypothermia, plasma epinephrine concentrations may increase as much as 10-fold over the prebypass concentrations; norepinephrine concentrations typically increase to a lesser extent (4-fold) (2,31,37,39), and deeper levels of hypothermia attenuate these (Table 11.1). In early studies, peak increases in both norepinephrine and epinephrine occurred when the heart and lungs were excluded from the circulation (38,39,40). However, norepinephrine and epinephrine concentrations peaked at different times. In a later study, patients undergoing cardiac surgery were randomly assigned to have CPB with mild (34°C) or moderate (28°C) hypothermia. With both bypass temperatures, peak norepinephrine concentrations were observed after release of the aortic crossclamp and rewarming, whereas peak epinephrine concentrations were observed at the target hypothermic temperature (42). A more recent study demonstrated a biphasic plasma norepinephrine concentration response to nonpulsatile CPB, with concentrations peaking at aortic declamping and again 2 to 4 hours after surgery. Epinephrine concentrations did not show this pattern, nor was it observed during pulsatile CPB (40,41). Neonates, infants, and young children, much like adults, demonstrate marked increases in catecholamine concentrations during CPB (2,41,43,44,45) (Fig 11.4).

Deeper planes of general anesthesia (whether accomplished with larger doses of synthetic opioids, addition of a propofol infusion, higher concentrations of volatile anesthetic vapors, or addition of neuraxial anesthesia) significantly reduce the catecholamine concentrations of patients undergoing coronary artery bypass surgery compared with patients less deeply anesthetized (46,47,48). Furthermore, in critically ill neonates undergoing correction of congenital heart disease, deeper planes of general anesthesia from large intravenous doses of sufentanil not only produced lower catecholamine concentrations in response to CPB, but also reduced mortality compared with lighter planes of general anesthesia using halothane and morphine (2) (Fig 11.4). Consistent with these observations regarding anesthetic depth, infusion of propofol during CPB (4 mg/kg/hr) resulted in markedly reduced concentrations of epinephrine and norepinephrine compared with a single bolus injection of diazepam 0.1 mg/kg (47). Addition of thoracic epidural anesthesia to a “high-dose” fentanyl or sufentanil general anesthetic significantly reduced catecholamine concentrations during and after CPB relative to concentrations measured without thoracic epidural anesthesia (49,50) (Fig. 11.5). Similarly, patients undergoing CPB after “high spinal” intrathecal blockade had reduced levels of catecholamines compared to a control group, despite neither group receiving a high-dose opioid general anesthetic (51) (Fig. 11.5).

The effect of pulsatile perfusion on catecholamine concentrations during CPB remains controversial (31,52). Although early studies demonstrated that catecholamine concentrations were increased during CPB whether or not pulsatile perfusion was used (31), a more recent study of elective coronary surgery patients showed significant reductions in epinephrine and norepinephrine concentrations with pulsatile (vs. nonpulsatile) perfusion (40,52) (Fig. 11.6).

Some increase in catecholamine concentrations during and after CPB may be unavoidable with current anesthetic and surgical techniques; nevertheless, deeper planes of general anesthesia (either with larger doses of opioids or greater concentrations of inhaled general anesthetics) or addition of conduction anesthesia to general anesthesia can limit the increases.

Increased secretion of cortisol is one of the central features of the metabolic stress response. In the classic studies by Hume et al. (53) of patients undergoing major (noncardiac) surgery, cortisol concentrations rose quickly to a maximum and then slowly returned to baseline 24 hours postoperatively. It is therefore not surprising that even “off-pump” coronary revascularization procedures are associated with increases in serum cortisol and other markers of the stress response despite the absence of the stress of CPB (54). It is apparent, however, that CPB modifies cortisol responses to surgery. Total plasma cortisol concentrations typically briefly decrease immediately upon initiation of bypass, likely as a consequence of hemodilution (55,56,57,58) (Fig. 11.7). During bypass, cortisol concentrations rise to values significantly above baseline (2,5,55,56,57,58,59,60). After CPB, patients exhibit markedly elevated concentrations of cortisol (both free and total) for more than 48 hours (58,59,60,61). In a recent study, cortisol increased 5-fold from baseline 2 hours after “on pump” coronary artery surgery versus a maximal 3-fold increase after “off pump” coronary artery surgery (60). Interestingly, the maximal increase occurred significantly earlier in the “off pump” group (4 hours after surgery) than in the “on pump” group (12 hours after surgery). The authors speculated that this finding resulted from differences in inflammatory processes between the two forms of coronary artery surgery. Cortisol levels slowly decreased thereafter in both groups.

Leptin, an adipocyte-derived hormone, is thought to moderate the acute systemic inflammatory response to CPB and surgery and to interact importantly with the hypothalamic-pituitary-adrenal axis. Leptin binds to receptors in the hypothalamus and is known to affect energy metabolism. It has structural similarities to cytokines, it affects immunity, and it may modulate stress responses. Leptin concentrations decrease with cardiac surgery and CPB. “On pump” coronary surgical cases showed a more pronounced decrease in leptin compared to “off pump” coronary surgeries. Both groups also showed a subsequent increase in leptin 24 hours postoperatively. Leptin levels correlated inversely with plasma cortisol levels (60,62,63). Children undergoing surgical repair of congenital heart diseases with CPB demonstrated similar findings, with leptin concentrations decreasing during CPB and increasing afterward to peak at 12 hours postoperatively. Leptin concentrations were elevated in critically ill patients; thus elevations in leptin concentrations may serve as a marker for systemic inflammatory response syndrome (64).

Tinnikov et al. (65) studied 14 children undergoing repair of ventricular septal defects with deep hypothermia and circulatory arrest, but without CPB. Maximal perioperative concentrations of cortisol and minimal perioperative concentrations of cortisol binding globulin were recorded at the first assessment after circulatory arrest. Thus, hypothermia and circulatory arrest initiate a cortisol-stress response even in the absence of extracorporeal perfusion.

Cortisol responses during bypass appear to be temperature-dependent. Taggart et al. (66) showed that the increase in cortisol concentration during CPB can be blunted by perfusion with blood at 20°C compared to 28°C. Peak CPB cortisol concentrations were decreased by deeper planes of anesthesia in both adults and children (2,58,59) (Fig. 11.8). Winterhalter et al. (48) showed that when continuous remifentanil (0.25 µg/kg/min) was compared to intermittent bolus dose fentanyl (2.6 ± 0.3 mg/kg total dose) corticotropin (adrenocorticotropic hormone, ACTH), cortisol, and vasopressin were all significantly decreased. This result could be secondary to the specific agent (remifentanil vs. fentanyl), variations in the depth of anesthesia, or the steady state produced by continuous anesthetic infusions. Stenseth et al. (49) found that, compared with high-dose fentanyl anesthesia alone, high-dose fentanyl anesthesia plus thoracic epidural anesthesia delayed the increase in cortisol concentrations during coronary artery surgery and reduced concentrations during bypass. Similarly, Moore et al. (50) found that thoracic epidural anesthesia combined with sufentanil 20 µg/kg was associated with markedly lower cortisol concentrations as compared to sufentanil anesthesia alone. On the other hand, spinal anesthesia failed to attenuate cortisol responses (compared to intravenous general anesthesia) in children undergoing correction of congenital heart defects (3).

CPB modifies ACTH responses in surgical patients. In the previously mentioned study by Hume et al. (53), non-CPB surgical patients showed no increase in cortisol concentrations after an injection of ACTH, indicating that adrenal secretion of cortisol was already maximal. Amado and Diago (67) observed a blunted response to corticotropin-releasing hormone during bypass, similar to responses seen in patients with hypothalamic corticotropin-releasing hormone deficiency. In contrast, an earlier study revealed that when patients undergoing extracorporeal perfusion received ACTH, cortisol concentrations increased (55).

Taylor et al. (68) measured a progressive fall in ACTH concentrations during bypass, with a subsequent increase 1 hour after pulsatile perfusion was restored. Raff et al. (57) showed that, although neither high-dose fentanyl anesthesia nor dexamethasone 40 mg alone blunted the increase in ACTH concentration in response to CPB, concurrent administration of both agents significantly reduced the ACTH concentration (Fig. 11.7). In a 2011 study, Debono and colleagues (69) demonstrated that at least 25% of patients undergoing coronary artery bypass grafting (CABG) who had normal cosyntropin stimulation tests prior to surgery developed increased ACTH concentrations and decreased responses to cosyntropin (an ACTH derivative used in diagnostic testing) postoperatively. The clinical importance of this relative cortisol deficiency is unclear as postoperative outcomes were comparable regardless of ACTH concentrations or cosyntropin responses.

Unlike some other hormones, cortisol and ACTH responses to CPB generally have not been influenced by pulsatile perfusion. To be sure, one study found that total plasma cortisol rose during pulsatile bypass but fell dramatically in patients undergoing nonpulsatile perfusion (56). In another study, patients with and without pulsatile perfusion showed initial increases in cortisol, ACTH, and aldosterone, followed by a gradual decline in concentrations of all three hormones during bypass and then a subsequent increase in all three hormones after bypass perfusion (70). After correction for the effect of hemodilution, there was no decrease in calculated free cortisol concentrations and a slight increase in adrenocorticotropic hormone concentrations, irrespective of pulsatile versus nonpulsatile perfusion. In children with either pulsatile or nonpulsatile perfusion, Pollock et al. (71) found large increases in cortisol and ACTH during CPB, followed by a slow decline toward baseline concentrations of both hormones over 24 hours with both techniques.

Although there is no unequivocal evidence for adrenocortical hypofunction during or after CPB, the inflammatory response initiated by the triad of blood contact with the foreign surfaces of the extracorporeal membrane, reperfusion injury, and endotoxemia may be attenuated by large doses of exogenous glucocorticoids (72). This inflammatory response triggers tissue injury in the heart, kidneys, hemostatic system, and especially the lung, which is the only organ exposed to the entire cardiac output (except during CPB). Early investigations have studied small numbers of cardiac surgery patients randomized to variable doses of different corticosteroids (most commonly 1 mg/kg dexamethasone or 30 mg/kg methylprednisolone) initiated at varying intervals between induction of anesthesia and the start of CPB (72). Overall, study results generally demonstrate an amelioration of the inflammatory response, with decreases in cytokine formation (tumor necrosis factor and the interleukin [IL]-1, -6, and -8) but inconsistent effects on C3a and elastase concentrations. Leukotrienes such as LTB4 are decreased in a dose-dependent fashion (72,73). IL-10, a cytokine with actions that are principally anti-inflammatory, demonstrates increased concentrations with steroid administration, supporting an anti-inflammatory effect (73). In addition, large doses of methylprednisolone can block upregulation of neutrophil integrin adhesion receptors, whereas dexamethasone decreases endothelial production of certain adhesion molecules (74). Clinically, glucocorticoid therapy may increase cardiac index and decrease systemic vascular resistance (75). Dietzman et al. (76) showed improvement in tissue perfusion and a decrease in peripheral vascular resistance when a large glucocorticoid dose was given just before CPB. Routine glucocorticoid supplementation has also been advocated as part of an accelerated recovery program (77), albeit without much supporting evidence. A small study has shown that cardiac surgery patients receiving glucocorticoids had shorter lengths of stay and improved quality of life as compared to patients not receiving glucocorticoids (78). The same research group found that steroid treatment reduced the duration of catecholamine support, length of intensive care unit (ICU) stay, and likelihood of postoperative atrial fibrillation (79). A 2013 meta-analysis that included 48 randomized controlled trials (RCTs) of patients receiving corticosteroids undergoing CPB failed to show any difference in the incidence of myocardial infarction, stroke, renal insufficiency, or death. The only outcome effect identified was a modest and heterogeneous decreased ICU and hospital length of stay in the groups receiving steroids (73).

In summary, current data nearly uniformly demonstrate large increases in cortisol and ACTH concentrations with initiation of CPB. These increases may be attenuated by deeper planes of general anesthesia or by addition of thoracic epidural (but apparently not spinal) anesthesia to general anesthesia as well as by continuous infusions of narcotic agents. Pulsatile perfusion does not appear to reduce these exaggerated responses. Moreover, it is not clear whether elevated corticosteroid concentrations during bypass are deleterious or beneficial.