Metabolic Issues in Liver Transplantation Robert E. Shangraw, MD, PhD

Patients who undergo orthotopic liver transplantation (OLT), almost by definition, have end-stage liver disease. One cannot examine the metabolic changes observed during OLT without insight into the profound effects of preexisting liver dysfunction. Because the liver plays a central role in regulation of whole-body metabolism, its disease leads to major alterations in glucose metabolism. Most patients presenting for OLT have chronic liver disease which has gradually progressed to the point where replacement surgery is indicated. The endocrine and metabolic changes associated with chronic cirrhosis are well described. A small subset of patients presents for OLT with fulminant liver failure which may or may not have been preceded by chronic liver disease. Evidence describing metabolic alterations of patients with fulminant disease, because of its unstable nature, is much less complete than that for patients with chronic disease. Many of the underlying endocrine and metabolic alterations are exacerbated by the OLT procedure itself. After OLT, full normalization of glucose metabolism is delayed and corrects in stages. Certain primary inherited diseases (‘‘inborn errors’’) of metabolism, which can be corrected by OLT, present unique issues that are beyond the scope of this review. Patients with end-stage liver disease exhibit disturbances in lipid, amino acid, and protein metabolism; and also poor clearance of drugs whose elimination or biotransformation depends on hepatic function. However, these subjects are also outside of the scope of the present review. This review focuses on the intertwined derangements of glucose metabolism, acid-base balance, and serum electrolyte concentrations that appear in patients before OLT, then describes changes which occur intraoperatively and persist in the postoperative setting. The endocrine environment that in part mediates the metabolic response is presented first. 1

2 ’

’

Shangraw

Preoperative Endocrine and Metabolic Environment in Cirrhosis Hormonal Profile

Patients with chronic end-stage liver disease exhibit a peculiar endocrine profile. Serum insulin concentration is elevated at least 2-fold, and glucagon concentration 5-fold, compared with healthy controls.1–3 Concurrently increased plasma C-peptide concentration indicates that hyperinsulinemia is due to augmented secretion more than delayed clearance by the diseased liver.2–4 Hyperglucagonemia is also due to increased secretion.1 Despite its usual role as a ‘‘counter-regulatory’’ hormone, glucagon is not an important mediator for ‘‘insulin resistance’’ of glucose metabolism associated with cirrhosis.5,6 Hyperglucagonemia does have cardiovascular effects, enhancing cardiac output via increased contractility and heart rate.7,8 Glucagon also vasodilates splanchnic vessels and renders them refractory to vasoconstriction mediated by norepinephrine, angiotensin, or vasopressin.8–10 Atrial natriuretic factor has an inconsistent pattern in cirrhosis, whereas increased plasma renin activity and hyperaldosteronism contribute to Na+ retention.11 However, the typical presentation of patients with end-stage cirrhosis is hyponatremia, indicating that retention of free water outstrips that of Na+. A factor underlying free water retention in cirrhosis is excessive vasopressin antidiuretic hormone secretion.12 Treatment of fluid and Na+ retention in cirrhosis, usually initiated early in the course of the disease, includes diuretics and a Na+ -restricted diet. However, the clinical response is usually incomplete and hyponatremia persists. Glucose and Lactate Metabolism

Plasma glucose concentration in end-stage cirrhosis is increased above that of healthy volunteers, and may or may not attain the threshold for diagnosis of diabetes mellitus.13–17 The response of cirrhotic patients to glucose challenge, either oral or intravenous, is an exaggerated hyperglycemic and hyperinsulinemic response, which again may trigger the diagnosis of frank diabetes mellitus.13–18 Concomitant hyperglycemia and hyperinsulinemia define the phenomenon of insulin resistance. Cirrhosis-induced insulin resistance is caused by impaired peripheral (eg, skeletal muscle) glucose disposal.19–21 Within muscle, the 2 principal defects are failures to synthesize glycogen and to oxidize glucose.17 Patients with chronic cirrhosis, as a rule, exhibit normal hepatic glucose production.22,23 Hypoglycemia only occurs in the setting of fulminant hepatic failure causing a profound decrease in hepatic glucose production. However, maintenance of

Metabolic Effects of OLT

’

3

hepatic glucose production in cirrhotics depends upon the marked hyperglucagonemia: When glucagon secretion is suppressed by somatostatin infusion, hepatic glucose production decreases precipitously and hypoglycemia can result.22 Carbon dioxide accounts for 60%, and lactate another 25%, of whole-body glycolytic glucose disposal, and these fractions are unchanged in patients with end-stage liver disease.2 Resting lactate flux in healthy subjects and patients with cirrhosis is 2 to 3 mmol/kg/min.2 Plasma lactate uptake, which at steady state equals its production rate, occurs by the liver (40% to 50%), kidney (30%), and skeletal muscle (20%).24–26 Kidney and muscle have limited capacity to take up lactate, but may assume a larger role in the setting of liver disease. Plasma lactate taken up by the liver is incorporated into glucose (70%) or oxidized to CO2 (30%).27 Glucose-lactate cycling (Cori cycle) maintains a relatively constant plasma glucose concentration over the cycle of fasting and feeding. Although fasting lactate flux is normal in patients with cirrhosis, the response to lactate challenge is impaired, because of slower hepatic clearance.28 Patients with cirrhosis also respond to glucose challenge in an abnormal fashion, producing more lactate per gram of glucose taken up, resulting in hyperlactatemia.3,18,29 For these reasons, and their susceptibility to infection (eg, subacute bacterial peritonitis), patients with cirrhosis are at risk for developing lactic acidosis.

Preoperative Acid-base Balance

Most patients with stable end-stage cirrhosis have neutral acid-base balance. They can, however, present with disturbances that include respiratory alkalosis, metabolic alkalosis, or metabolic acidosis. Unless a patient has been treated with a drug that decreases respiratory drive, or is profoundly fatigued, respiratory acidosis is unlikely. The most common acid-base disturbance is primary respiratory alkalosis,30 mediated by a central nervous system mechanism. Metabolic alkalosis is usually iatrogenic owing to diuretic therapy with furosemide or a thiazide, most often accompanied by hypokalemia.31 Less often the underlying cause is protracted vomiting with loss of H+. Infrequently the patient may present with metabolic acidosis. The most common scenario producing metabolic acidosis is fulminant hepatic failure and consequent hepatorenal syndrome, although even in acute liver failure metabolic alkalosis is more common than metabolic acidosis.32 Two operant mechanisms may produce metabolic acidosis: (1) failure of the liver to clear lactate, resulting in lactic acidosis, and (2) inability of the dysfunctional kidney to retain bicarbonate. Although metabolic acidosis is an uncommon acid-base abnormality in cirrhosis, it is the most ominous. It may indicate urgent transplantation for fulminant hepatic

4

’

Shangraw

failure, or dialysis as a temporizing measure to stabilize the patient before and possibly during OLT.31 Preoperative Electrolyte Abnormalities

Patients with end-stage cirrhosis exhibit hyponatremia due to free water retention that exceeds the concomitant sodium retention. Not only is severe hyponatremia itself a threat to the patient’s well being, but too rapid an increase in plasma Na+ concentration, by administration of normal saline or sodium bicarbonate, puts the cirrhotic patient at risk of central pontine myelinolysis.33 Serum K+ concentration may be high, low, or within normal limits.34 Hypokalemia is often secondary to furosemide or thiazide diuretic therapy.31 Hyperkalemia can be caused by (1) potassiumsparing diuretics such as spironolactone or amiloride, (2) metabolic acidosis that stimulates exchange of intracellular potassium for extracellular H+ , or (3) renal failure from a primary cause or secondary to hepatorenal syndrome. Serum K+ concentration in end-stage liver disease does not reflect total body K+ content, which is usually decreased (‘‘K+ depletion’’) compared with that in healthy subjects.34 Endogenous insulin is important to prevent hyperkalemia.35–37 Compared with healthy subjects, cirrhotic patients challenged with KCl exhibit a higher peak serum K+ concentration along with an exaggerated hyperinsulinemia to dispose of the K+, indicating a rightward shift of the dose-response curve and another expression of insulin resistance.38 The maximal hypokalemic response to insulin, however, is preserved in end-stage cirrhosis.39 Serum Ca2+ concentration is usually within normal limits, but can be decreased by a substantial plasma transfusion due to its chelation by citrate in the transfusate.40–42 Similarly, circulating Mg2+ concentration can be decreased,43 most often associated with poor nutrition, but is usually within normal limits in patients receiving good preoperative care. ’

Intraoperative and Postoperative Metabolic Changes With OLT Hormonal Milieu

Changes in intraoperative in serum insulin concentration are shown in Figure 1.39 Preoperative hyperinsulinemia increases during the dissection stage, and the rate of rise accelerates during the anhepatic stage, due in part to loss of insulin clearance by the liver. Peak insulin concentration occurs at the end of the anhepatic stage and early after graft reperfusion. Hyperinsulinemia subsequently moderates but

Metabolic Effects of OLT

’

5

Figure 1. Serum insulin concentration during OLT (n = 10). Time points: 0 = Before anesthesia induction; I-5 = 5 minutes before skin incision; II-60 = 60 minutes before anhepatic; II + 30, II + 60 = 30 and 60 minutes anhepatic, respectively; III + 15, III + 30, III + 60, III + 120, III + 180 = 15, 30, 60, 120, and 180 minutes after portal vein unclamping, respectively. Data are means ± SE. *P<0.05 versus I-5 value, wP<0.05 versus healthy volunteers. Data from Shangraw and Hexem.39

continues above baseline throughout the remainder of the procedure and postoperative course. Hyperglucagonemia sustains throughout the procedure, moderated only by dilution from transfusion and later by the hyperinsulinemia associated with reperfusion.39 Serum glucagon concentration can be suppressed by high-dose insulin infusion during the dissection and anhepatic stages, but nevertheless remains far increased above that in healthy subjects39 (Fig. 2). Hyperinsulinemia and hyperglucagonemia continue for years after OLT.44–46

Figure 2. Serum glucagon concentration during OLT (n = 10/group). Time points are identical to Figure 1. Open circles (*) are OLT alone, filled circles () are during insulin infusion 500 mU/m2/min from anesthesia induction. Data are means ± SE. *P<0.05 versus I-5 value, wwP<0.05 versus healthy volunteers. From Shangraw and Hexem.39

6

’

Shangraw

Glucose and Lactate Metabolism

Glucose A progressive hyperglycemia occurs during the dissection and anhepatic phases of OLT despite a concurrent rise in serum insulin concentration39,47–50 (Fig. 3). Plasma glucose concentration during this time can be controlled by insulin infusion.39 Hypoglycemia in adults is rare during the anhepatic stage, although somewhat more likely in the setting of fulminant hepatic failure without preceding chronic disease. The absence of hypoglycemia during the anhepatic stage is due to peripheral insulin resistance and continued gluconeogenesis by extrahepatic tissues.51 Underlying causes for the progressive hyperglycemia include (1) baseline insulin resistance; (2) administered glucocorticoid (methylprednisolone) exacerbating insulin resistance52,53; (3) the stress response with increased secretion of counterregulatory hormones (norepinephrine, epinephrine, cortisol); (4) therapeutic infusion of catecholamines such as dopamine or epinephrine; and (5) direct infusion of glucose in transfused blood products. Portal vein unclamping is followed by an immediate stepwise increase in plasma glucose concentration.39,47–50 The sudden hyperglycemia is due to glucose release by the graft liver, and occurs even when the patient has previously had good intraoperative glycemic control with insulin infusion.39 Hepatic glucose release, a unique stress response by the liver, is due to glycogenolysis begun during organ storage and accelerating during rewarming and after perfusion.54 Some anesthesiologists argue that an exaggerated and protracted hyperglycemia after reperfusion portends poor graft function. The patient with frank diabetes mellitus before surgery is also most likely to demonstrate exaggerated hyperglycemia after reperfusion. Many patients who did not need insulin for preoperative glycemic management require insulin

Figure 3. Plasma glucose concentration during OLT (n = 10/group). Time points are identical to Figure 1. Open circles (*) are OLT alone, filled circles () are during insulin infusion 500 mU/m2/min from anesthesia induction. Data are means ± SE. *P<0.05 versus I-5 value, wP<0.05 versus II + 60 value, wwP<0.05 versus healthy volunteers. From Shangraw and Hexem.39

Metabolic Effects of OLT

’

7

treatment after graft reperfusion. Insulin infusion required for glycemic control after reperfusion must often be continued postoperatively. Although there are theoretical reasons to expect cirrhosis-induced insulin resistance to disappear after transplantation, glycemic management often still requires insulin therapy. This is because (1) the hyperglycemic stress response to surgery continues into the postoperative setting, and may in fact be greater once anesthesia is discontinued; and (2) immunosuppression, whether by prednisone, cyclosporine, or tacrolimus, induces glucose intolerance and hyperglycemia itself.53,55–57 Cyclosporine and tacrolimus both inhibit endogenous insulin secretion but do not greatly compromise peripheral responsiveness to insulin.55–58 Compared with before OLT, the liver transplant recipient is more likely to require insulin therapy, or more aggressive insulin therapy,in the intermediate postoperative term. Glucose tolerance improves after 5 months, and insulin-dependent glucose metabolism normalizes by 26 months, most likely related to decreased glucocorticoid dosing.46 Lactate At anesthesia induction, plasma lactate concentration is usually within normal limits, although it may be higher than that in patients without liver disease about to undergo surgery (Fig. 4).2,47 Throughout the dissection stage there is a progressive increase in plasma lactic acid concentration.2,47,59 The rate of accumulation for lactate accelerates during the anhepatic stage. This occurs because impairment of lactate clearance is exacerbated when the liver is excluded from the circulation. Tissue ischemia does not seem to be a factor in most cases,59 although there may be substantial blood loss and transfusion. Blood in the transfusion reservoir may have a lactate

Figure 4. Plasma lactate concentration during OLT (n = 33/group). Time points are identical to Figure 1. Open circles (*) are OLT alone, filled circles () are with dichloroacetate (DCA) 40 mg/kg at anesthesia induction and repeated at 4 hours. Data are means ± SE. *P<0.001 versus OLT alone. From Shangraw et al.47

8

’

Shangraw

concentration of 5 to 7 mmol/L,47 with higher concentrations possible if the blood is near its outdate or has prolonged warming in the reservoir where erythrocytes produce lactate as their principal metabolic product. A further incremental rise in plasma lactate concentration occurs immediately after graft reperfusion. By 1 hour, plasma lactate concentration stabilizes, then slowly decreases as the new liver begins to metabolize the lactate. At the end of surgery, plasma lactate concentration is still elevated above baseline but is decreased from peak concentration.47,60 Plasma lactate concentration decreases in the early postoperative course, returning to baseline or normal limits within the first postoperative day. If the graft liver is functioning poorly, plasma lactate concentration may continue to increase in the postoperative setting.61 Dichloroacetate (DCA) is an experimental agent that activates mitochondrial pyruvate dehydrogenase and decreases lactic acid formation via (1) enhancing oxidation of pyruvate in equilibrium with lactate, and (2) inhibiting glycolysis.2,62 DCA slows intraoperative accumulation of lactic acid by 50% (Fig. 4).2,47,63 DCA does not reduce hyperglycemia, indicating that the excess glucose output does not result from accelerated lactate-derived gluconeogenesis.47

Acid-base Balance

Most patients for OLT exhibit normal acid-base balance at the outset of surgery. This may be due to premedication and/or controlled ventilation after induction of anesthesia eliminating a primary respiratory alkalosis.30 During OLT, there is a biphasic acid-base disturbance, characterized by initial metabolic acidosis then protracted metabolic alkalosis.47,48,50,59,60 The severity of metabolic acidosis parallels the accumulation of lactic acid (Figs. 4, and 5), indicating that the major etiology is lactic acidosis.47 Moreover, attenuation of lactate accumulation by DCA stabilizes acid-base balance.47,60 Loss of base equivalents starts during the dissection stage and accelerates during the anhepatic stage.47,48 The nadir for arterial pH and base deficit occurs in the first several minutes after graft reperfusion (Fig. 5). The main consequence of metabolic acidosis is impaired cardiovascular function. In dogs, lactic acidosis inhibits myocardial contractility such that when compared with a control pH 7.4, an arterial pH of 7.3, 7.2, and 7.1 leads to a decrease in cardiac output of 14%, 20%, and 29%, respectively, because of impaired stroke volume and contractility.64 Contractile dysfunction occurs secondary to a reversible impairment of myocardial metabolism.65,66 Acidemia also causes hypotension by decreasing vascular tone,67 which is already hyporesponsive in patients with cirrhosis.8–10 There is some evidence that acidemia exacerbates coagulopathy.68

Metabolic Effects of OLT

’

9

Figure 5. Acid-base parameters during OLT (n = 33/group). Top panel shows arterial pH and bottom panel shows arterial base excess. Time points are identical to Figure 1. Open circles (*) are OLT alone, filled circles () are with dichloroacetate (DCA) 40 mg/kg at anesthesia induction and repeated at 4 hours. Data are means ± SE. *P<0.04 versus OLT alone, wP<0.02 versus basal value. From Shangraw et al.47

Several options are available to treat intraoperative metabolic acidosis. First is respiratory compensation, by increasing the minute ventilation, decreasing PaCO2, and returning arterial pH towards 7.4. Beyond a critical base deficit, however, the most common treatment is NaHCO3. Administering bicarbonate effectively replaces base equivalents, but its high Na+ concentration (1000 mEq/L) is problematic in hyponatremic cirrhotic patients, as copious use increases the risk of central pontine myelinolysis.33,69–71 NaHCO3 can also, with rapid administration, produce a paradoxical myocardial depression most likely secondary to worsened intracellular acidosis.72–74 A large mass of administered bicarbonate exacerbates the magnitude of metabolic alkalosis after graft reperfusion. When copious replacement of base equivalents is anticipated, such as in renal failure, a reasonable alternative is tris-hydroxymethyl aminomethane, which although a weaker base, has a much lower Na+ content and, with a pK of 7.8, is a more effective buffer than NaHCO3 in the physiological pH range.75

10

’

Shangraw

Further, it lessens rather than exacerbates intracellular acidosis.75 Trishydroxymethyl aminomethane side effects of respiratory depression, hypoglycemia, hypokalemia and peripheral vein irritation are largely irrelevant in the patient undergoing OLT.75 A patient with preoperative metabolic acidosis may need intraoperative dialysis or continuous hemofiltration. DCA preempts lactic acidosis, specifically decreasing lactic acid formation.47,60,62 It is not a buffer, nor does it provide base equivalents to generically counter metabolic acidosis. Its sole utility is for treatment of lactic acidosis.2,62 Unfortunately DCA is not commercially available as a pharmaceutical agent. The second half of the biphasic intraoperative pH response occurs once the graft liver begins to clear H+ along with lactate and citrate. Acid-base balance returns towards neutral, then a metabolic alkalosis ensues by 2 hours after portal vein unclamping47,48,50,59,60 (Fig. 5). Most patients exhibit metabolic alkalosis at the completion of surgery, which can persist for several days postoperatively. Its etiology is multifactorial, initially caused as the liver clears excess lactate and citrate from plasma.61,76 Two H+ equivalents are stoichiometrically consumed for each lactate, and 3 for each citrate, metabolized. Citrate load correlates with the severity of postoperative metabolic alkalosis.76 However, plasma concentrations of lactate and citrate return to normal by the first postoperative day,42,47 indicating that other mechanism(s) must come into play beyond 24 hours. Urea synthesis, a process that consumes HCO3, could underlie metabolic alkalosis if it was impaired. However, the 10-fold stimulation of urea synthesis exhibited by patients recovering from OLT rules this out as a mechanism causing postoperative metabolic alkalosis.63 Diuretic treatment that many patients receive may be an important mechanism maintaining metabolism-induced alkalemia.31 Compensation for postoperative metabolic alkalosis is decreased respiratory drive and hypoventilation. The patient becomes at risk for hypoxemia and atelectasis, and is less likely to satisfy criteria for extubation. This protracts mechanical ventilation and intensive care unit stay. Possible treatments for severe metabolic alkalosis include acetezolamide (diamox) to interfere with bicarbonate reabsorption or, in extreme cases, HCl infusion to consume bicarbonate.31

Potassium Balance

Patients are usually eukalemic at anesthesia induction, but can alternatively exhibit either hypokalemia or hyperkalemia. Although intraoperative serum K+ concentration is stable in most patients39,49 (Fig. 6), it can increase dramatically at any time during the dissection or anhepatic stages. This can be due to (1) cellular K+ efflux in exchange

Metabolic Effects of OLT

’

11

Figure 6. Serum potassium concentration in pulmonary artery during OLT (n = 9). Time points are: pre = Before anesthesia induction; A = anesthetized, before skin incision; B = dissection phase; C = anhepatic phase; D = portal vein unclamping; E = biliary reconstruction; F = skin closure. Inset shows first 3 minutes after portal vein unclamping. Data are means ± SE. *P<0.05 versus anhepatic, or **P<0.01 versus anhepatic. From Carmichael et al49 with permission.

for H+ uptake in a milieu of metabolic acidosis77; (2) K+ intolerance secondary to cirrhosis38; and (3) exogenous K+ infusion due to a high K+ content in the blood product transfusion reservoir. Erythrocytes lyse and expel K+ into the fluid as they senesce, and the transfusate K+ concentration can reach 20 mEq/L. Rapid infusion of fluid with high K+ concentration can quickly lead to life-threatening hyperkalemia at almost any time during the operation. Hyperglucagonemia may play a role in the K+ intolerance of cirrhosis.38,78 Immediately after reperfusion there is a brief period (3 to 5 min) of hyperkalemia (Fig. 6), after which hypokalemia most often ensues.49 The main source of the transient excess K+ is the graft liver.79 The graft has been bathed in preservative (‘‘UW solution’’) with a high K+ content which requires careful flushing. The liver can also naturally release a large quantity of intracellular K+ when stressed.80 In addition, marked acidemia immediately after reperfusion can mobilize K+ from the large intracellular pool in all tissues.77 The etiology of subsequent postreperfusion hypokalemia is hyperinsulinemia induced by concomitant hyperglycemia (Fig. 3). Insulin stimulates the graft liver to avidly take up K+ , almost immediately restoring a normal whole-body maximal hypokalemic response to insulin.39

12

’

Shangraw

Treatment of hyperkalemia is essential to prevent cardiac arrest. Insulin is the most rapid and consistent agent to decrease serum K+ concentration, far exceeding the speed and magnitude of other mechanisms in vivo.81,82 The molecular mechanism by which insulin decreases serum K+ concentration is independent of and faster than its action on glucose metabolism83–86 (Fig. 7). After insulin binds to its membrane receptor, cellular K+ uptake ultimately involves activation of Na-K ATPase. In liver, there is intermediate activation of sodiumhydrogen exchanger (NHE1), and insulin-mediated K+ uptake is antagonized by glucagon or amiloride.78,87 NHE1 is not involved in the skeletal muscle response.88,89 Potassium uptake in vivo occurs within seconds after insulin administration, contrasting with the glucose

Figure 7. Schematic diagram of insulin-mediated K + uptake by liver and skeletal muscle. Details and supporting references are in text.

Metabolic Effects of OLT

’

13

response that requires 30 to 60 minutes. The majority (70%) of insulinstimulated K+ uptake occurs across the splanchnic bed.39,86 Other tissues (eg, muscle) eventually take up more but the brisk response occurs via the liver. This tissue distribution differs from that of glucose uptake, for which 60% of whole-body uptake occurs by skeletal muscle. The maximal hypokalemic response to insulin in patients undergoing OLT does not differ from that by healthy subjects.39,86 Although insulin can decrease serum K+ during the anhepatic stage, the response is only 20% to 30% of the magnitude observed in the presence of a liver.39,90 It is reasonable to ensure a serum K+ r4.5 mEq/L, with insulin if necessary, before clamping the portal vein. Because most insulinmediated K+ uptake is trapped in the native liver, and removed at hepatectomy, there is no risk of rebound hyperkalemia as reported with the glucose-insulin-K+ technique during cardiac surgery.91 Marked insulin resistance of glucose metabolism makes it unnecessary to coadminister glucose with the insulin. It is better to monitor plasma glucose concentration at 30-minute intervals to determine whether glucose administration is indicated. Maximal hypokalemic response occurs at a serum insulin concentration about 1/5 of that needed for maximal hypoglycemic effect.86 In a normal human without ongoing blood loss, this translates to a bolus 20 units versus 60 units regular insulin, respectively. Hypokalemia after reperfusion is corrected by slow titrated administration of KCl. KCl requirement may be increased if an insulin infusion has been initiated for glycemic control.

Citrate Metabolism, Calcium and Magnesium Balance

Little is known about the preoperative kinetics of citrate metabolism in cirrhosis. Plasma citrate concentration in patients at the time of anesthesia induction, however, is within normal limits42 (Fig. 8). Plasma citrate concentration rises modestly during the dissection phase, then exhibits a steep rise during the anhepatic stage, such that the plasma concentration may exceed baseline by 10-fold by the end of the anhepatic stage.42,92,93 Rapid clearance of citrate by the graft liver begins almost immediately after portal vein unclamping, and plasma citrate concentration typically returns to near-normal concentration by the completion of surgery. The magnitude of rise correlates with the quantity of plasma transfused, and is largely due to exogenous rather than endogenous citrate. However, plasma citrate concentration increases even when no plasma is transfused.92 The intraoperative increase in plasma citrate concentration is pathologic, reflecting limited ability of the liver to clear citrate from the plasma. The 2 main intraoperative consequences of hypercitratemia are hypocalcemia and hypomagnesemia secondary to chelation of the free

14

’

Shangraw

Figure 8. Plasma citrate and serum calcium concentrations during OLT (n = 91). Time points are: A1 = After anesthesia induction; A2 = dissection phase  5 minute; A3 = 5 minute before portal vein clamping; B1 = anhepatic  5 minute; B2 = anhepatic  60 minutes; C1 = reperfusion  5 minute; C2 = reperfusion  60 minutes; C3 = end surgery; D1 = postop  24 hours; D2 = postop  48 hours; D3 = postop  72 hours. Data are means ± SD. Statistics: (a) = P<0.05 versus A1 value, (b) = P<0.01 versus A1 value, (c) = P<0.001 versus A1 value. Reprinted from Clin Biochem, Vol 28, Diaz J et al.73 Correlation among ionized calcium, citrate, and total calcium levels during hepatic transplantation, pp. 315–317. Copyright 1995, with permission from Elsevier.

ions by citrate.40,41,92–94 Decreases in ionized electrolyte concentrations are much more pronounced than that for the total electrolyte concentrations. The effect of citrate on Ca2+ , which has its nadir in the late anhepatic stage, is superimposed in Figure 8. Hypocalcemia can provoke cardiovascular instability due to impaired cardiac contractility and diminished vascular tone.40,41,95 It should be corrected in the setting of poor hemodynamic function. On the other hand, it is reasonable to withhold Ca2+ administration absent hemodynamic instability, because rapid clearance of citrate after reperfusion is accompanied by release of Ca2+ as citrate is metabolized. Rebound hypercalcemia occurs in the first 2 hours after reperfusion. Excessive Ca2+ increases the risk of pancreatic injury in many different clinical scenarios.96,97 Intraoperative changes in serum ionized Mg2+ concentration parallel those of Ca2+ , with a hypomagnesemic nadir at the anhepatic stage, then recovering after reperfusion.92,93 Consequences of hypomagnesemia during OLT are cardiac irritability and exacerbation of clotting dysfunction, both of which are improved by Mg2+ replacement.94,98,99 Caution should be exercised in Mg2+ replacement during the anhepatic stage, guided by serum Mg2+ monitoring, to avoid rebound hypermagnesemia after reperfusion with consequent

Metabolic Effects of OLT

’

15

hyporeflexia, diminished mental status and prolonged neuromuscular blockade. The major late consequence of hypercitratemia is exacerbation of metabolic alkalosis, because each citrate metabolized is accompanied by consumption of 3 H+ equivalents. An epidemiologic review determined that the strongest predictor for postoperative metabolic alkalosis was the number of intraoperative units of plasma transfused.76 ’

Summary and Future Directions

Anesthetic management for OLT is complex because of preexisting metabolic derangements of cirrhosis, intraoperative removal of the central metabolic regulatory organ, and introduction of a metabolically active organ previously subjected to prolonged ischemia. Preoperative metabolic disturbances are intrinsic to the indication for surgery, limiting how well the patient can be ‘‘optimized’’ for surgery. Nevertheless, the intraoperative metabolic problems presented in this review can be anticipated, for which understanding the underlying pathophysiology facilitates intraoperative and postoperative management. As we learn more about the cellular and molecular mechanisms affecting metabolism in cirrhosis, and possible interventions, perioperative management will become more elegant. It has come a long way since the advent of clinical OLT surgery nearly 50 years ago. Supported in part by PHS Grants DK-19525 and 5 M01 RR000334, and a Clinical Scientist Research Award from the International Anesthesia Research Society.

’

References

1. Sherwin RS, Fisher M, Bessoff J, et al. Hyperglucagonemia in cirrhosis: altered secretion and sensitivity to glucagon. Gastroenterology. 1978;74:1224–1228. 2. Shangraw RE, Jahoor F. Mechanism of dichloroacetate-induced hypolactatemia in humans with or without cirrhosis. Metabolism. 2004;53:1087–1094. 3. Johnston DG, Alberti KG. Hormonal control of ketone body metabolism in the normal and diabetic state. Clin Endocrinol Metab. 1982;11:329–361. 4. Kruszynska YT, Home PD, McIntyre N. Relationship between insulin sensitivity, insulin secretion and glucose tolerance in cirrhosis. Hepatology. 1991;14:103–111. 5. Greco AV, Rebuzzi AG, Altomonte L, et al. Glucose, insulin and somatostatin infusion for the determination of insulin resistance in liver cirrhosis. Horm Metab Res. 1979;11:547–549. 6. Del Prato S, Castellino P, Simonson DC, et al. Hyperglucagonemia and insulinmediated glucose metabolism. J Clin Invest. 1987;79:547–556. 7. Abel FL. Action of glucagon on canine left ventricular performance and coronary hemodynamics. Circ Shock. 1983;11:45–58. 8. Silva G, Navasa M, Bosch J, et al. Hemodynamic effects of glucagon in portal hypertension. Hepatology. 1990;11:668–673.

16

’

Shangraw

9. Richardson PD, Withrington PG. The inhibition by glucagon of the vasoconstrictor actions of noradrenaline, angiotensin and vasopressin on the hepatic arterial vascular bed of the dog. Br J Pharmacol. 1976;57:93–102. 10. Benoit JN, Granger DN. Splanchnic hemodynamics in chronic portal hypertension. Semin Liver Dis. 1986;6:287–298. 11. Gadano A, Moreau R, Heller J, et al. Relation between severity of liver disease and renal oxygen consumption in patients with cirrhosis. Gut. 1999;45:117–121. 12. Moller S, Bendtsen F, Henriksen JH. Vasoactive substances in the circulatory dysfunction of cirrhosis. Scand J Clin Lab Invest. 2001;61:421–429. 13. Megyesi C, Samols E, Marks V. Glucose tolerance and diabetes in chronic liver disease. Lancet. 1967;2:1051–1056. 14. Collins JR, Crofford OB. Glucose intolerance and insulin resistance in patients with liver disease. Arch Intern Med. 1969;124:142–148. 15. Collins JR, Lacy WW, Stiel JN, et al. Glucose intolerance and insulin resistance in patients with liver disease. II. A study of etiologic factors and evaluation of insulin actions. Arch Intern Med. 1970;126:608–614. 16. Stone BG, Van Thiel DH. Diabetes mellitus and the liver. Semin Liver Dis. 1985;5: 8–28. 17. Petrides AS, Schulze-Berge D, Vogt C, et al. Glucose resistance contributes to diabetes mellitus in cirrhosis. Hepatology. 1993;18:284–291. 18. Kruszynska YT, Meyer-Alber A, Darakhshan F, et al. Metabolic handling of orally administered glucose in cirrhosis. J Clin Invest. 1993;91:1057–1066. 19. Proietto J, Alford FP, Dudley FJ. The mechanism of the carbohydrate intolerance of cirrhosis. J Clin Endocrinol Metab. 1980;51:1030–1036. 20. Cavallo-Perin P, Cassader M, Bozzo C, et al. Mechanism of insulin resistance in human liver cirrhosis. Evidence of a combined receptor and postreceptor defect. J Clin Invest. 1985;75:1659–1665. 21. Shmueli E, Walker M, Alberti G, et al. Normal splanchnic but impaired peripheral insulin-stimulated glucose uptake in cirrhosis. Hepatology. 1993;18:86–95. 22. Keller U, Sonnenberg GE, Burckhardt D, et al. Evidence for an augmented glucagon dependence of hepatic glucose production in cirrhosis of the liver. J Clin Endocrinol Metab. 1982;54:961–968. 23. Shangraw RE, Jahoor F. Effect of liver disease and transplantation on urea synthesis in humans: relationship to acid-base status. Am J Physiol. 1999;276:G1145–G1152. 24. Yudkin J, Cohen RD. The contribution of the kidney to the removal of a lactic acid load under normal and acidotic conditions in the conscious rat. Clin Sci Mol Med. 1975;48:121–131. 25. Buchalter SE, Crain MR, Kreisberg R. Regulation of lactate metabolism in vivo. Diabetes Metab Rev. 1989;5:379–391. 26. Ferrannini E, Natali A, Brandi LS, et al. Metabolic and thermogenic effects of lactate infusion in humans. Am J Physiol. 1993;265:E504–E512. 27. Magnusson I, Schumann WC, Bartsch GE, et al. Noninvasive tracing of Krebs cycle metabolism in liver. J Biol Chem. 1991;266:6975–6984. 28. Woll PJ, Record CO. Lactate elimination in man: effects of lactate concentration and hepatic dysfunction. Eur J Clin Invest. 1979;9:397–404. 29. Leatherdale BA, Chase RA, Rogers J, et al. Forearm glucose uptake in cirrhosis and its relationship to glucose tolerance. Clin Sci (London). 1980;59:191–198. 30. Moreau R, Hadengue A, Soupison T, et al. Arterial and mixed venous acid-base status in patients with cirrhosis. Influence of liver failure. Liver. 1993;13:20–24. 31. Shangraw RE. Acid-base balance. In: Miller RD, ed. Anesthesia. 5th ed. Philadelphia: Churchill Livingstone; 2000:1390–1413. 32. Record CO, Iles RA, Cohen RD, et al. Acid-base and metabolic disturbances in fulminant hepatic failure. Gut. 1975;16:144–149.

Metabolic Effects of OLT

’

17

33. Sterns RH. The treatment of hyponatremia: first, do no harm. Am J Med. 1990;88:557–560. 34. Casey TH, Summerskill WH, Orvis AL. Body and serum potassium in liver disease. I. relationship to hepatic function and associated factors. Gastroenterology. 1965;48: 198–207. 35. Santeusanio F, Faloona GR, Knochel JP, et al. Evidence for a role of endogenous insulin and glucagon in the regulation of potassium homeostasis. J Lab Clin Med. 1973;81:809–817. 36. Pettit GW, Vick RL. Contribution of pancreatic insulin to extrarenal potassium homeostasis: a two-compartment model. Am J Physiol. 1974;226:319–324. 37. DeFronzo RA, Sherwin RS, Dillingham M, et al. Influence of basal insulin and glucagon secretion on potassium and sodium metabolism. Studies with somatostatin in normal dogs and in normal and diabetic human beings. J Clin Invest. 1978;61: 472–479. 38. Decaux G, Soupart A, Cauchie P, et al. Potassium homeostasis in liver cirrhosis. Arch Intern Med. 1988;148:547–548. 39. Shangraw RE, Hexem JG. Glucose and potassium metabolic responses to insulin during liver transplantation. Liver Transplant Surg. 1996;2:443–454. 40. Marquez J, Martin D, Virji MA, et al. Cardiovascular depression secondary to ionic hypocalcemia during hepatic transplantation in humans. Anesthesiology. 1986;65: 457–461. 41. Martin TJ, Kang Y, Robertson KM, et al. Ionization and hemodynamic effects of calcium chloride and calcium gluconate in the absence of hepatic function. Anesthesiology. 1990;73:62–65. 42. Diaz J, Acosta F, Parrilla P, et al. Correlation among ionized calcium, citrate, and total calcium levels during hepatic transplantation. Clin Biochem. 1995;28:315–317. 43. Rocchi E, Borella P, Borghi A, et al. Zinc and magnesium in liver cirrhosis. Eur J Clin Invest. 1994;24:149–155. 44. Francavilla A, Polimeno L, Van Thiel DH, et al. Pancreatic hormones and amino acid levels following liver transplantation. Hepatology. 1987;7:918–924. 45. Shangraw RE, Jahoor F. Lipolysis and lipid oxidation in cirrhosis and after liver transplantation. Am J Physiol Gastrointest Liver Physiol. 2000;278:G967–G973. 46. Luzi L, Perseghin G, Regalia E, et al. Metabolic effects of liver transplantation in cirrhotic patients. J Clin Invest. 1997;99:692–700. 47. Shangraw RE, Winter R, Hromco J, et al. Amelioration of lactic acidosis with dichloroacetate during liver transplantation in humans. Anesthesiology. 1994;81: 1127–1138. 48. Farman JV, Lines JG, Williams RS, et al. Liver transplantation in man. Anaesthetic and biochemical management. Anaesthesia. 1974;29:17–32. 49. Carmichael FJ, Lindop MJ, Farman JV. Anesthesia for hepatic transplantation: cardiovascular and metabolic alterations and their management. Anesth Analg. 1985;64:108–116. 50. Zadrobilek E, Mauritz W, Feil W, et al. Hemodynamic and metabolic changes during orthotopic liver transplantation. Anaesthesist. 1988;37:672–679. 51. Lauritsen TL, Grunnet N, Rasmussen A, et al. The effect of hepatectomy on glucose homeostasis in pig and in man. J Hepatol. 2002;36:99–104. 52. Rizza RA, Mandarino LJ, Gerich JE. Cortisol-induced insulin resistance in man: impaired suppression of glucose production and stimulation of glucose utilization due to a postreceptor defect of insulin action. J Clin Endocrinol Metab. 1982;54: 131–138. 53. Pagano G, Cavallo-Perin P, Cassader M, et al. An in vivo and in vitro study of the mechanism of prednisone-induced insulin resistance in healthy subjects. J Clin Invest. 1983;72:1814–1820.

18

’

Shangraw

54. Nowak G, Ungerstedt J, Wernerman J, et al. Metabolic changes in the liver graft monitored continuously with microdialysis during liver transplantation in a pig model. Liver Transplant. 2002;8:424–432. 55. Robertson RP, Franklin G, Nelson L. Intravenous glucose tolerance and pancreatic islet beta-cell function in patients with multiple sclerosis during 2-yr treatment with cyclosporin. Diabetes. 1989;38:58–64. 56. Dresner LS, Andersen DK, Kahng KU, et al. Effects of cyclosporine on glucose metabolism. Surgery. 1989;106:163–169. 57. Van Duijnhoven EM, Christiaans MH, Boots JM, et al. Glucose metabolism in the first 3 years after renal transplantation in patients receiving tacrolimus versus cyclosporine-based immunosuppression. J Am Soc Nephrol. 2002;13: 213–220. 58. Uchizono Y, Iwase M, Nakamura U, et al. Tacrolimus impairment of insulin secretion in isolated rat islets occurs at multiple distal sites in stimulus-secretion coupling. Endocrinology. 2004;145:2264–2272. 59. Steib A, Freys G, Gohard R, et al. Tissue oxygenation during liver transplantation. Crit Care Med. 1992;20:977–983. 60. Shangraw RE, Robinson ST. Oxygen metabolism during liver transplantation: the effect of dichloroacetate. Anesth Analg. 1997;85:746–752. 61. Svensson KL, Persson H, Henriksson BA, et al. Whole body gas exchange: amino acid and lactate clearance as indicators of initial and early allograft viability in liver transplantation. Surgery. 1989;105:472–480. 62. Shangraw RE, Rabkin JM, Lopaschuk GD. Hepatic pyruvate dehydrogenase activity in humans: effect of cirrhosis, transplantation, and dichloroacetate. Am J Physiol. 1998;274:G569–G577. 63. Shangraw RE, Fisher DM. Pharmacokinetics and pharmacodynamics of dichloroacetate in patients with cirrhosis. Clin Pharmacol Ther. 1999;66:380–-390. 64. Teplinsky K, O’Toole M, Olman M, et al. Effect of lactic acidosis on canine hemodynamics and left ventricular function. Am. J Physiol. 1990;258: H1193–H1199. 65. Zhou HZ, Malhotra D, Shapiro JI. Contractile dysfunction during metabolic acidosis: role of impaired energy metabolism. Am J Physiol. 1991;261:H1481–H1486. 66. Zahler R, Barrett E, Majumdar S, et al. Lactic acidosis: effect of treatment on intracellular pH and energetics in living rat heart. Am J Physiol. 1992;262: H1572–H1578. 67. Yartsev VN, Karachentseva OV, Dvoretsky DP. Effect of pH changes on reactivity of rat mesenteric artery segments at different magnitude of stretch. Acta Physiol Scand. 2002;174:1–7. 68. Cosgriff N, Moore EE, Sauaia A, et al. Predicting life-threatening coagulopathy in the massively transfused trauma patient: hypothermia and acidoses revisited. J Trauma. 1997;42:857–861. 69. Wszolek ZK, McComb RD, Pfeiffer RF, et al. Pontine and extrapontine myelinolysis following liver transplantation. Relationship to serum sodium. Transplantation. 1989; 48:1006–1012. 70. Estol CJ, Faris AA, Martinez AJ, et al. Central pontine myelinolysis after liver transplantation. Neurology. 1989;39:493–498. 71. Abbasoglu O, Goldstein RM, Vodapally MS, et al. Liver transplantation in hyponatremic patients with emphasis on central pontine myelinolysis. Clin Transplant. 1998;12:263–269. 72. Ayus JC, Krothapalli RK. Effect of bicarbonate administration on cardiac function. Am J Med. 1989;87:5–6. 73. Ritter JM, Doktor HS, Benjamin N. Paradoxical effect of bicarbonate on cytoplasmic pH. Lancet. 1990;335:1243–1246.

Metabolic Effects of OLT

’

19

74. Thompson CH, Syme PD, Williams EM, et al. Effect of bicarbonate administration on skeletal muscle intracellular pH in the rat: implications for acute administration of bicarbonate in man. Clin Sci (London). 1992;82:559–564. 75. Kraut JA, Kurtz I. Use of base in the treatment of severe acidemic states. Am J Kidney Dis. 2001;38:703–727. 76. Contreras G, Garces G, Reich J, et al. Predictors of alkalosis after liver transplantation. Am J Kidney Dis. 2002;40:517–524. 77. Nakasuji M, Bookallil MJ. Pathophysiological mechanisms of postrevascularization hyperkalemia in orthotopic liver transplantation. Anesth Analg. 2000;91: 1351–1355. 78. Hallbrucker C, vom Dahl S, Lang F, et al. Inhibition of hepatic proteolysis by insulin. Role of hormone-induced alterations of the cellular K+ balance. Eur J Biochem. 1991;199:467–474. 79. Chapin JW, Wood RP, Hurlbert BJ, et al. Sources of increased serum potassium following reperfusion of liver allografts. Transplant Proc. 1987;19:51–53. 80. Shoemaker WC. The action of epinephrine and other hormones associated with the stress response on potassium movements with special reference to the development of postoperative depletion states. Rev Surg. 1968;25:9–24. 81. Allon M, Shanklin N. Effect of bicarbonate administration on plasma potassium in dialysis patients: interactions with insulin and albuterol. Am J Kidney Dis. 1996; 28:508–514. 82. Greenberg A. Hyperkalemia: treatment options. Semin Nephrol. 1998;18:46–57. 83. Zierler KL. Effect of insulin on potassium efflux from rat muscle in the presence and absence of glucose. Am J Physiol. 1960;198:1066–1070. 84. Andres R, Baltzan MA, Cader G, et al. Effect of insulin on carbohydrate metabolism and on potassium in the forearm of man. J Clin Invest. 1962;41:108–115. 85. Burton SD, Mondon CE, Ishida T. Dissociation of potassium and glucose efflux in isolated perfused rat liver. Am J Physiol. 1967;212:261–266. 86. DeFronzo RA, Felig P, Ferrannini E, et al. Effect of graded doses of insulin on splanchnic and peripheral potassium metabolism in man. Am J Physiol. 1980; 238:E421–E427. 87. Lynch CJ, Mader AC, McCall KM, et al. Okadaic acid stimulates ouabain-sensitive 86Rb(+ )-uptake and phosphorylation of the Na+ /K(+ )-ATPase alpha-subunit in rat hepatocytes. FEBS Lett. 1994;355:157–162. 88. Weil E, Sasson S, Gutman Y. Mechanism of insulin-induced activation of Na(+ )K(+ )-ATPase in isolated rat soleus muscle. Am J Physiol. 1991;261:C224–C230. 89. Ragolia L, Cherpalis B, Srinivasan M, et al. Role of serine/threonine protein phosphatases in insulin regulation of Na+ /K+ -ATPase activity in cultured rat skeletal muscle cells. J Biol Chem. 1997;272:23653–23658. 90. De Wolf A, Frenette L, Kang Y, et al. Insulin decreases the serum potassium concentration during the anhepatic stage of liver transplantation. Anesthesiology. 1993;78:677–682. 91. Bohrer H, Fleischer F, Krier C. Hyperkalemic cardiac arrest after cardiac surgery following high-dose glucose-insulin-potassium infusion for inotropic support. Anesthesiology. 1988;69:949–953. 92. Kulpmann WR, Rademacher E, Bornscheuer A. Ionized magnesium concentration during liver transplantation, resection of the liver and cardiac surgery. Scand J Clin Lab Invest Suppl. 1996;224:235–243. 93. Scott VL, De Wolf AM, Kang Y, et al. Ionized hypomagnesemia in patients undergoing orthotopic liver transplantation: a complication of citrate intoxication. Liver Transplant Surg. 1996;2:343–347. 94. Bennett MW, Webster NR, Sadek SA. Alterations in plasma magnesium concentrations during liver transplantation. Transplantation. 1993;56:859–861.

20

’

Shangraw

95. Drop LJ. Ionized calcium, the heart, and hemodynamic function. Anesth Analg. 1985;64:432–451. 96. Fernandez-del Castillo C, Harringer W, Warshaw AL, et al. Risk factors for pancreatic cellular injury after cardiopulmonary bypass. N Engl J Med. 1991;325:382–387. 97. Niederau C, Luthen R, Klonowski-Stumpe H, et al. The role of calcium in pancreatitis. Hepatogastroenterology. 1999;46:2723–2730. 98. Ranasinghe DN, Mallett SV. Hypomagnesaemia, cardiac arrhythmias and orthotopic liver transplantation. Anaesthesia. 1994;49:403–405. 99. Choi JH, Lee J, Park CM. Magnesium therapy improves thromboelastographic findings before liver transplantation: a preliminary study. Can J Anaesth. 2005;52: 156–159.