Preoperative and Postoperative Hepatic Dysfunctions

Introduction

It is estimated that up to 10% of cirrhotic patients will undergo surgery in the final 2 years of their lives. This leads to the need for a careful review of preoperative and postoperative liver dysfunction in this group of patients. In this chapter we will review many aspects of what is known with regard to how the normal and diseased liver reacts to operative procedures.

The liver is the major site for the metabolism of many drugs and is responsible for the synthesis of most serum proteins and the removal of endogenous and exogenous toxins. Therefore postoperative liver dysfunction may slow the recovery of a patient who has undergone surgery. Because of the liver's role in drug metabolism, it is susceptible to injury by a variety of xenobiotics. Alterations in hepatic blood flow may also affect liver function, especially in the patient with underlying chronic liver disease. Therefore it is not unexpected that liver test result abnormalities are noted frequently in patients after surgery. However, clinical jaundice is rare (<1%) in patients with normal livers, and its development should prompt a thorough evaluation of its cause. Before reviewing the causes of postoperative liver dysfunction, we will briefly discuss the evaluation of the patient with abnormal liver test results found during preoperative testing. If the surgery is critical to the patient, the presence of preoperative liver dysfunction should have no effect on the decision to operate. However, if the surgery is elective, the finding of abnormal liver test results preoperatively should prompt an evaluation as to their cause, and an estimate of liver function and reserve should be made.

Preoperative Liver Dysfunction

The frequency of unsuspected liver disease is approximately 1 in 700 of otherwise healthy surgical candidates, making this a common clinical problem. Of most concern to surgeons and anesthesiologists are elevations of the serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), and bilirubin levels, and disorders of coagulation. Elevations of alkaline phosphatase or γ-glutamyltranspeptidase levels are of little clinical significance and should not prompt an evaluation unless they are associated with other clinical findings (see Chapter 7 ). The significance of a low serum albumin level alone is difficult to interpret because of the multiple causes of a fall in the concentration of this protein (see Chapter 7 ). It is useful to divide patients with preoperative liver dysfunction into three groups: (1) asymptomatic with normal physical examination findings; (2) symptomatic; and (3) physical and biochemical findings of chronic liver disease.

Asymptomatic patients with normal physical examination findings and abnormal liver test results (increased AST and ALT levels) are encountered frequently (9.8% of adults in the United States), raising concerns that they have underlying liver disease that may increase the risk of surgery. The abnormal liver test results most likely reflect a subacute or chronic form of liver injury, resulting from viral hepatitis, drugs, and/or alcohol, or especially fatty liver. We lack information as to whether this group of patients is at any increased risk from surgery, but it is the author's opinion that if the serum bilirubin, albumin, and clotting test results are normal and the increase in the levels of aminotransferases is mild (twofold to fivefold), there is little, if any, increase in surgical risk. Deciding to perform surgery and ensuring its successful completion in this type of patient should not be an end to the evaluation of the liver disease. After surgery the liver test findings need to be monitored; if they are persistently abnormal, a thorough evaluation should be performed.

Greater increases in the levels of aminotransferases are more problematic only because they suggest more significant hepatic injury. Anesthetic agents may adversely affect hepatic function because of decreases in splanchnic blood flow and thereby oxygen delivery to the liver. Early studies suggested that this may be a real clinical concern when an increase in operative mortality was observed in patients with acute hepatitis. However, in other studies no increase in mortality has been found when surgery was performed in patients with coincidental acute viral hepatitis. All of these studies suffer from being anecdotal and reporting on small numbers of patients. In addition, many patients were jaundiced, suggesting the presence of significant liver injury. Despite these uncertainties, if the surgery is elective, a delay in the operation is the most conservative approach. Liver tests should be performed 2 weeks to 3 weeks later, and if the abnormalities persist for several months the patient should be evaluated completely. If it is decided to perform an elective operation when AST and ALT levels is elevated more than fivefold but a normal albumin level, a normal prothrombin time, and a normal bilirubin level is noted, the risk to the patient probably remains small. However, if the liver test results worsen postoperatively, defining whether the surgery or the initial illness is at fault will be difficult. There is no evidence that the presence of liver disease increases the risk of developing anesthetic-induced hepatitis (see Chapter 56 ); there is therefore no need to alter the choice of anesthetic agent used for the surgery.

Isolated elevations in the serum bilirubin level are usually due to Gilbert syndrome (see Chapter 58 ); this syndrome does not increase the risk of surgery. If the elevations of bilirubin level are seen in association with elevations of AST and ALT levels or the alkaline phosphatase level, the cause of the liver injury needs to be determined before any elective surgery is performed. Isolated prolongation of the prothrombin time is an unusual manifestation of liver disease but suggests the presence of cirrhosis or a severe acute or subacute injury.

The symptomatic patient with elevated liver test concentrations is of greater concern when elective surgery is planned. The presence of symptoms (nausea, vomiting) in association with elevated liver test concentrations suggests that the patient is developing an acute illness that may worsen before it abates. Therefore if the patient is subjected to surgery and the test results worsen, it will be very difficult to determine whether the patient is suffering from a preexisting condition or a complication of surgery (i.e., anesthetic-induced hepatitis). In addition, there is uncertainty as to whether or not acute viral hepatitis increases the risk of surgery ; therefore elective surgery should be postponed until the hepatitis has resolved.

A special subset of the symptomatic patient is represented by patients with markedly elevated levels of aminotransferases. This group is varied as to cause of these elevations: the causes range from ischemic hepatitis or cardiac dysfunction, drugs, liver trauma, and cancer metastases to the liver to rhabdomyolysis. A review demonstrated that the overall mortality of patients with a serum AST level greater than 3000 IU/L was 55% and that in ischemic hepatitis patients it was 75% compared with 33% for all other causes. This group should clearly be excluded from consideration for any elective surgery.

The presence of clinical (i.e., splenomegaly, spiders, palmar erythema, ascites) and biochemical (i.e., low albumin level, prolonged prothrombin time) evidence of chronic liver disease is of greatest concern when one is planning elective surgery. The risk of the surgery is determined by how well the liver is functioning and the presence or absence of symptoms. For example, patients with chronic hepatitis appear to have an increased surgical risk if they are symptomatic. Elective surgical procedures in patients with evidence of chronic liver disease should be delayed until the cause of the liver disease has been determined and the severity of the injury has been fully assessed.

The most common cause of chronic liver disease in the Western world is alcoholism. Alcoholic liver disease can manifest itself as fatty liver, alcoholic hepatitis, cirrhosis, or a combination of these conditions (see Chapter 23 ). The risk associated with surgery in patients with fatty liver appears to be small. It is not uncommon to find unsuspected cirrhosis in obese patients at the time of bariatric surgery. Among 125 patients with cirrhosis detected at surgery, Brolin et al. were able to proceed with their planned bariatric surgery in 74% of the patients; the result was no intraoperative deaths and a 4% mortality rate. In a larger review, patients with Child class A/B cirrhosis were at greater risk of complications (21.3%), liver decompensation (6.55%), early mortality (1.6%), and late mortality (2.45%) compared with those without cirrhosis. The risk also did not appear to be influenced by the type of gastric bypass. The presence of hepatitis C without cirrhosis may increase the risk of complications following orthopedic surgery without a change in mortality. Alcoholic hepatitis can be present as an asymptomatic illness or may be associated with jaundice and liver failure. Elective surgery in a patient with decompensated liver disease resulting from any cause is ill-advised. Even in patients with better preserved liver function, elective surgery in patients with acute symptomatic alcoholic hepatitis is associated with increased morbidity and mortality and should not be performed. The effect of asymptomatic alcoholic hepatitis on surgical mortality has been studied best in patients undergoing portosystemic shunt surgery. The presence of large amounts of alcoholic hyaline in a liver biopsy specimen indicated a high likelihood of death in some series. In another series, the 1-year survival rate of patients after the insertion of a portosystemic shunt was 70% to 74% in the absence of alcoholic hyaline and only 10% in those with alcoholic hyaline in a liver biopsy. In contrast, other investigators have found no correlation between the presence of alcoholic hyaline and survival rates. Despite these uncertainties, elective surgery in a patient with clinical features of alcoholic hepatitis should be avoided. It is very difficult, however, to determine whether an alcoholic patient has a fatty liver or a more serious lesion (i.e., alcoholic hepatitis) on the basis of liver test results alone. Therefore the most conservative approach in a chronic alcoholic patient who requires elective surgery and who has abnormal liver test results is either to perform a liver biopsy to define the nature of the liver injury or a period of abstinence (2 months to 3 months) to allow the acute injury to resolve before surgery.

There is little question that the presence of cirrhosis increases the risk of performing surgery, especially if it is an intraabdominal operation. Published mortality rates range from 0% to 100% depending on the type of surgery and the severity of liver disease. These differences in survival rates reflect the variability in the clinical state of the patients reported in the different studies. The risk of performing surgery is defined best by the clinical severity of the cirrhosis (e.g., Child classification, Model for End-Stage Liver Disease [MELD] score). Patients with good hepatic function, no ascites, and a good nutritional state (Child class A) do well with surgery, whereas those with jaundice, low serum albumin levels, ascites, and muscle wasting have a high operative mortality and postoperative morbidity. Fig. 57-1 shows the relationship between the MELD score and postoperative mortality at 30 days and 90 days. Although there is no clear inflection point, when the MELD score exceeds 10, the expected mortality exceeds 10%, a meaningful number when one is discussing elective surgery. For MELD scores greater than 20, the expected mortality exceeds 50%. A single-point increase in the MELD score was associated with a 14% to 15% increase in the risk of death at 30 days and 90 days respectively. Elective surgery can be performed in patients with decompensated liver disease but the surgeon must have experience in the treatment of this type of patient to minimize the risks of surgery. Emergent surgery appears to carry a high risk of in-hospital death for certain procedures as compared with elective operations. The presence of portal hypertension also appeared to increase the surgical risk.

A difficult problem that many surgeons face is performing a cholecystectomy in a patient with cirrhosis. Mortality rates of 7.5% to 25.5% and morbidity rates of 4.8% to 25% have been reported. In addition, many patients require transfusions, especially if they have decompensated liver disease. Factors associated with increases in morbidity and mortality include low serum albumin levels, coagulopathy, higher Child-Pugh score, and more bleeding during surgery. A subtotal cholecystectomy used to be suggested for patients with cirrhosis and portal hypertension. In the past, open cholecystectomy was recommended for patients with cirrhosis, but studies since the 1990s suggest that in Child class A and class B patients, elective laparoscopic cholecystectomy is preferred because there is less bleeding, fewer wound infections, and less time in the hospital than with open cholecystectomy. The complication rate after urgent cholecystectomy is significantly higher (36%) than that after elective laparoscopic cholecystectomy (16%) in cirrhotic patients, suggesting that, if possible, stabilization of the patient with medical management followed by elective surgery would be preferable. Given the high rates of mortality and morbidity in patients with advanced cirrhosis undergoing cholecystectomy and the increased difficulty associated with liver transplant in patients with a previous cholecystectomy, the author believes that the presence of recurrent cholecystitis in a patient with advanced liver disease should be considered an indication for liver transplant while conservative management is being performed. This may require an appeal to be filed with the United Network for Organ Sharing for U.S. patients to allow them to have a higher-priority MELD score. In the acutely ill patient, placement of percutaneous cholecystostomy tubes may be attempted but this may be difficult to perform in patients with significant ascites.

Complication rates are also quite high for cirrhotic patients with abdominal surgical procedures other than just cholecystectomy. The overall mortality rate for urgent abdominal surgery has been reported to be 45% to 57% versus 10% to 18% with elective surgeries. Important predictive factors associated with mortality in some studies include the presence of ascites ( p = 0.006), encephalopathy ( p = 0.002), and coagulopathy ( p = 0.021). The mortality rates by the Child classification likewise demonstrate a progressive increased risk as liver disease worsens and are reported as 10% for patients with class A cirrhosis, 30% to 31% for patients with class B cirrhosis, and 76% to 82% for patients with class C cirrhosis. The causes of death from gastrointestinal tract surgical procedures in a series by Mansour et al. were coagulopathy or sepsis in 81% of cases. A 1984 report classified 87% of deaths as being due to multisystem organ failure. These data suggest that the surgery should be converted from an emergent procedure to an elective one, wherever possible, but this may depend on the situation. It remains to be seen if the new MELD scores will be more predictive of operative morality for these procedures than is the Child classification.

Given that coagulopathy plays a major role in postoperative abdominal surgery complications seen in cirrhotic patients, it is interesting to speculate what the future role of recombinant factor VIIa (rFVIIa) may be for these patients. Data on the use of rFVIIa have been increasingly seen in the literature, and there are scattered reports of use of this medication for limited abdominal procedures, such as diagnostic laparoscopy and some other radiologic and surgical procedures. It is unclear whether this effect is transient in nature, and if surgery is performed what impact this would have on overall postoperative bleeding complications and patient survival rates. Use of rFVIIa has been shown to be ineffective in controlling variceal bleeding, and its use during intraabdominal operations clearly has not been of any benefit.

Cardiac surgery in more advanced cirrhotic patients likewise has a reported high complication rate. Little mortality was noted in Child class A patients who underwent cardiac surgery but mortality was reported to be 42% to 50% in Child class B patients for similar surgical procedures and 100% in patients with Child class C cirrhosis. Higher MELD scores were also predictive of a worse outcome following cardiac surgery. Probable contributors to the increased mortality include liver perfusion injuries related to cardiopulmonary bypass, as well as bypass aggravation of the coagulopathy, which may already be present in these patients. New advances in the use of cardiac stents and valvuloplasty techniques may help these patients avoid thoracotomy. For example, transcatheter aortic value replacement has been successfully performed in patients with advanced liver disease.