Surgical Procedures That Increase the Risk of Postoperative Liver Failure

In addition to the severity of preoperative hepatic dysfunction, the actual surgical procedure performed may be the most important risk factor for postoperative liver failure.^[120] Under normal circumstances, the liver is able to adapt to changes in perfusion pressure by altering either

Figure 55-6 Effect of the number of risk factors on the perioperative complication rate in cirrhotic patients undergoing surgery. Risk factors included the Child-Pugh score, the presence of ascites, a diagnosis of cirrhosis other than primary biliary cirrhosis, serum creatinine concentration, a diagnosis of chronic obstructive pulmonary disease, preoperative infection, preoperative upper gastrointestinal bleeding, higher American Society of Anesthesiologists physical status score, intraoperative hypotension, and higher surgical severity score. (From Ziser A, Plevak DJ, Wiesner RH, et al: Morbidity and mortality in cirrhotic patients undergoing anesthesia and surgery. Anesthesiology 90:42–53, 1999.)

In addition to surgery on the biliary tract, stomach, and colon, hepatic resection for HCC is a known risk factor for the development of liver failure in patients with preoperative hepatic dysfunction. Most patients with HCC have liver dysfunction associated with either chronic hepatitis or cirrhosis.^{[106] [122]} The diminished functional hepatic reserve of these patients may decrease the amount of tissue that can be resected without compromising viable parenchyma and causing liver failure, the most common cause of death after this surgery. Wu and coworkers^[123] retrospectively analyzed cirrhotic patients undergoing hepatic resection for HCC based on their degree of hepatic dysfunction. By limiting the amount of tissue resection in patients with significant hepatic dysfunction (i.e., poorer Child-Pugh class), the authors noted 5-year survival rates comparable to those in patients with less severe forms of cirrhosis who underwent more extensive resections. Perioperative hemorrhage is a common occurrence in cirrhotic patients undergoing resection for HCC because of contributing factors such as portal hypertension, coagulation abnormalities, and highly vascular adhesions in those with previous abdominal surgery. Preoperative evaluation of patients with HCC by indocyanine green 15-minute retention^{[123] [124]} or direct measurement of the hepatic venous pressure gradient^[125] can be beneficial in predicting postoperative outcomes in cirrhotic patients undergoing hepatic resection for HCC.

2219

Cardiothoracic surgery appears to be associated with a high mortality rate in patients with preexisting liver disease.^{[115] [126] [127]} Hepatic dysfunction is often exacerbated by cardiopulmonary bypass (CPB), although the precise mechanism of this dysfunction has not been clearly elucidated. In a canine model of hypothermic nonpulsatile CPB, Koizumi and colleagues^[128] noted a decrease in PBF and HABF without a concomitant change in hepatic oxygen metabolism. Normothermic CPB was associated with a similar decline in PBF, but HABF was maintained. Increasing the dose of fentanyl from 10 to 50 µg/kg/hr significantly suppressed HABF and impaired hepatic oxygen metabolism in both normothermic and hypothermic CPB, thus suggesting the possibility of fentanyl-mediated peripheral venous pooling and subsequent lowering of cardiac output. More recently, Okano and coworkers^[129] assessed hepatosplanchnic oxygenation in 25 patients with no history of hepatic dysfunction who were undergoing elective coronary artery bypass grafting with either normothermic (>35°C) or hypothermic (<32°C) CPB. Hepatic venous desaturation and functional impairment of hepatic sinusoidal endothelial cells occurred in both groups, but hepatocellular dysfunction was not observed.

In addition to the possible effects of hepatic artery and portal venous perfusion, other potential determinants of hepatic dysfunction after CPB include hypotension, low cardiac output syndrome, hypoxemia, microembolism or macroembolism, cytokine and oxygen-free radical formation, and the influence of vasoactive and anesthetic drugs. Small retrospective studies have shown 25% to 30% mortality rates in Child-Pugh class A and B patients undergoing nonemergency cardiac surgery.^{[126] [127]} Although more deaths occurred in the Child-Pugh class B group, significant morbidity, particularly in the form of postoperative hemorrhage and infection, was seen in the Child-Pugh class A patients. Finally, noncardiac, thoracic surgery has been suggested as a risk factor for postoperative mortality in patients with cirrhosis, although the basis for this association remains speculative. ^[115]