Liver Disease

What are the risks of giving anesthesia to patients with acute liver disease who require emergency surgery? What are the risks of giving anesthesia to patients with chronic impairment of liver function? What can be done to minimize these risks? Although one might think that the experiences gained from providing anesthesia for liver transplantation would answer many of these questions, there is a substantial difference between optimizing cardiovascular function to meet the needs of a new liver (e.g., supply of nutrients) and maintaining liver function in a diseased liver. Because hepatic function and physiology are discussed in Chapter 55 , we will mention only that the liver performs many functions: it synthesizes substances (e.g., proteins, clotting factors), detoxifies the body of both drugs and the products of normal human metabolism, excretes waste products, and stores and supplies energy. Tests of liver function assess synthesis (cholesterol levels, prothrombin time [PT], albumin levels), cellular integrity (aspartate aminotransferase [AST], alanine aminotransferase [ALT], lactate dehydrogenase, alkaline phosphatase), the liver's ability to detoxify the body (e.g., ammonia, direct bilirubin, or lidocaine levels), and the liver's ability to excrete certain substances (sulfobromophthalein retention, total bilirubin levels).

In examining the effects of anesthesia (with or without surgery) on liver function and measures to reduce risk in patients with preexisting liver disease, investigators have often looked at one or more of the aforementioned tests or, more commonly, at major end points of morbidity (jaundice) or mortality. The evidence can be summarized as follows: the last data we have (too old now, but it is the last data we have) indicate that without anesthesia or surgery, approximately 1 in 700 to 800 patients who are otherwise healthy and scheduled for surgery will have abnormal preoperative results of liver function tests; of these patients, jaundice will develop in 1 in 3 (also see Chapter 25 for these studies).^{[819] [820]} In our preliminary study, every patient with abnormal liver function tests has been symptomatic preoperatively.^[351] Therefore, the risk may be less than described (see Chapter 25 ).^[821]

All anesthetics tested (general, narcotic-nitrous oxide, and regional) have caused transient abnormalities in liver function test results. These abnormalities were magnified by upper intra-abdominal surgery and occurred regardless of preexisting liver disease.^{[773] [822] [823] [824] [825] [826] [827] [828] [829]} Patients whose preoperative liver function tests are abnormal will obviously have a higher incidence of abnormal results on postoperative liver function tests.^{[773] [822] [823] [824] [826] [827]} Lacking in the literature are investigations that studied patients with compromised hepatic function to determine how to decrease the risk of surgery and anesthesia. Also lacking are comparisons of outcome after various perioperative strategies for managing different causes of hepatic dysfunction (i.e., viral [or even one virus versus another virus], toxic, or drug induced). In addition to preexisting hepatic disease and the operative site, hypokalemia, hypotension, sepsis, and the need for blood transfusion all contribute to postoperative hepatic dysfunction. Thus, anesthesia and surgery probably exacerbate hepatic disease and thereby clearly increase morbidity and mortality.^{[773] [822] [823] [826] [827]} Only one study, on alcoholic liver disease, has implied otherwise.^[773]

The main goal of the anesthesiologist is to avoid making the hepatic disease (with perhaps its metabolic and CNS toxicity) worse and increasing the chance of renal failure, coma, and death.^[830] Studies of portal hypertension have shown that mortality can be 50% when preoperative serum albumin concentrations are lower than 3 g/dL, preoperative serum bilirubin levels are greater than 3 mg/dL, and ascites and encephalopathy are present; mortality can be even higher when the PT is grossly abnormal. By contrast, mortality decreases to 10% when preoperative serum albumin levels are 3 to 3.5 g/dL, preoperative serum bilirubin levels are 2 to 3 mg/dL, PT is normal, and encephalopathy is absent. The risks of anesthetizing

1111

In the late 1970s and 1980s it was debated whether halothane should be given to patients with hepatitis or biliary tract disease. The National Halothane Study did not find halothane to cause massive hepatic necrosis—or any other hepatic abnormality—more frequently than any other anesthetic^[699] and, in fact, demonstrated its safety in biliary tract surgery. Prospective studies have been contradictory at best regarding whether repeat exposure to halothane within a short time elevates liver enzyme levels to a greater degree than other anesthetics do.^{[835] [836] [837] [838]} Because the incidence of hepatitis attributable only to halothane would be very low, very large groups would have to be studied. The incidence of postoperative jaundice and hepatitis not attributable to halothane is much greater than the incidence of hepatitis attributable only to halothane, and the absence of differentiating pathologic features makes halothane-induced hepatitis difficult to exclude as a possibility. Newer tests imply an immunologic basis to the continuation of either hepatitis or liver dysfunction after the initial mechanism (e.g., if hypoxia begets the injury, the immune mechanism continues it) and are supported by resolution of a prolonged case of halothane hepatitis after treatment.^[839] Although other cases that have resolved after steroid therapy have been presented, those cases did not, in fact, have a long enough period before the beginning of their resolution when the disease was stable to conclude that steroids made a difference and that this was not the natural course of the disease. One case is inconclusive but certainly implies that a controlled trial should be instituted. This subject is perhaps even more important now that several forms of viral hepatitis (acute and chronic) are treatable; left untreated, hepatitis C evolves to chronic type C hepatitis in 50% of patients and to cirrhosis in 20% of those in whom chronic hepatitis develops. Treatment with interferon-alfa (leukocyte) appears to reduce the risk of these consequences substantially.^{[753] [754]}

More sensitive tests of hepatocellular dysfunction after anesthesia and surgery, such as glutathione-S-transferase, reveal impairment more frequently.^[840] One is left to conclude that both metabolic and immunologic mechanisms can contribute to hepatocellular dysfunction after anesthesia and surgery. Are these related to anesthesia or surgery? Internists almost always point to anesthesia and must be educated to shed light on their bias, although this need is less with better diagnostic tests^{[753] [755]} and with the possibility of treatment of some forms of hepatitis.^{[753] [754]}

Thus, the strong bias of internists makes anesthetic-induced hepatitis ("halothane hepatitis") a popular diagnosis despite the higher incidence of viral and drug hepatitis.^{[773] [838] [839] [840] [841] [842] [843] [844] [845] [846] [847] [848] [849] [850] [851] [852] [853]} Another factor producing possible bias is the fact that animal models of halothane hepatitis incorporate liver hypoxia or pretreatment with polyvinyl chlorides, conditions that can themselves adversely affect liver function. No irrefutable evidence has implicated halothane as being better or worse than any other anesthetic for patients with preexisting liver disease. However, should liver disease worsen postoperatively, the tendency is to blame halothane (even when it has not been in the operating room for 20 years). Anesthesia is certainly less likely than hypoxia, trauma, viral hepatitis, drug-induced hepatitis, or sepsis to cause serious hepatic injury.^{[819] [852]} No one has yet determined whether a time limit exists within which repeat exposure to an anesthetic is more dangerous than exposure to various anesthetics or after which repeat anesthesia is as safe as a first exposure to an anesthetic. The common consistent feature of many forms of chronic liver disease is a vulnerability to decreased oxygen delivery to the hepatic artery. This vulnerability occurs because of increased resistance to portal vein blood flow. Many stimuli, including surgical palpation, appear to decrease oxygen delivery to the hepatic artery. Thus, anesthesia and surgery may affect hepatic well-being after surgery in a patient with an already compromised liver.

What should physicians do to avoid contracting hepatitis or giving it to patients? It is cost-effective for anesthesia personnel to receive the appropriate vaccines and take appropriate precautions.^{[821] [853] [854] [855] [856]} What should a physician with hepatitis B do to prevent infecting patients? That subject is discussed in detail elsewhere.^[857]

Liver disease severe enough to affect hepatic synthesis can impair the detoxification of many drugs,^{[839] [840] [852] [858] [859] [860] [861]} including muscle relaxants^{[858] [859] [860] [861]} (also see Chapter 13 ), and can disturb coagulation (also see Chapter 48 ) but may or may not affect anesthetic requirements greatly.^{[862] [863]} Administration of fresh frozen plasma may be needed to correct coagulation disorders.