How Does Critical Illness Alter the Liver?

Liver alterations are common in critical illness. These changes can be attributed to diverse factors, including systemic inflammation and poor perfusion but also to drugs and parenteral nutrition. Damage ranges from self-limited abnormalities in liver chemistry to fulminant organ failure. In addition, critical illness elicits profound changes in the concentrations of acute-phase proteins, plasma peptides synthesized by the liver as part of the response to a “danger” signal. Although some of these proteins help to control damage, the functions of others are obscure. Critical illness also induced a reprogramming of metabolic function. Finally, there is a severity-dependent disruption of phase I and II biotransformation and bile transport (i.e., excretory failure with important implications for pharmacotherapy in the intensive care unit [ICU]). The hepatobiliary excretory machinery appears exceptionally sensitive to inflammation. As a result, impaired excretion occurs in the absence of traditional markers of (ischemic) liver cell death, such as serum transaminases. Importantly, these changes—acute-phase protein elaboration, altered metabolism, and disrupted biotransformation—occur in parallel. With increasing severity of liver impairment, other functions, most notably synthesis of coagulatory proteins and glucose homeostasis, fail and impaired clearance of toxic compounds affects other organs, as in the case of hepatic encephalopathy. Although fulminant liver failure in previously healthy patients is rare, a deterioration of a preexisting liver disease is more common. The extent of the dysfunction is often underestimated, especially in surgical patients, and is associated with substantial morbidity and mortality. Episodes of decompensated cirrhosis are manifested as variceal bleeding, renal insufficiency, and encephalopathy.

Mechanisms and Manifestation of Liver Dysfunction

The liver is a highly perfused organ, and there are complex mechanisms to regulate liver microcirculatory blood flow. Under pathophysiologic conditions, these regulatory mechanisms can become ineffective and impaired. Together with macrocirculatory changes, an altered microcirculation may lead to profound changes in critically ill patients and may even reduce effective sinusoidal blood flow. Globally reduced (e.g., in hemorrhagic shock, right heart failure, mechanical ventilation) or redistributed (i.e., sepsis, anaphylaxis, endocrinopathies) blood flow with shunting are mechanisms for ischemic damage. Shunts can be intrahepatic or extrahepatic. In patients with chronic liver disease and portal hypertension, an especially large amount of portal blood can be redirected via extrahepatic shunts. Conversely, increased liver blood flow in acute hepatitis has also been described.

Ischemic damage to the liver may accompany low flow states (i.e., shock) or reflect congestion as a consequence of right heart failure. When severe, ischemia characteristically leads to cell death in the pericentral region of the hepatic acinus, which is reflected in an increase in serum levels of glutamate dehydrogenase. Hypoxic hepatitis leading to centrolobular hepatocellular necrosis is associated with a rapid increase in serum aminotransferases (aspartate and alanine aminotransferase) with levels up to 20 times the upper limit of normal.

Impaired excretory function is a frequent manifestation of critical illness. Development of jaundice as a complication of severe trauma or life-threatening disease was only observed with the widespread establishment of ICUs in the 1960s. This finding may well be related to prototypical therapeutic ICU interventions such as multiple transfusions, parenteral nutrition, and potentially hepatotoxic medications. In a large Austrian multicentric cohort, an early increase in plasma bilirubin (>2 mg/dL), noted in approximately 10% of critically ill intensive care patients, was a strong independent risk factor for subsequent mortality.

Hepatocellular excretory dysfunction in the critically ill might result from altered blood flow or transmembrane transport. However, it is more frequently associated with systemic inflammation than with ischemic damage. Sepsis accounts for approximately 20% of jaundice cases. This number is surpassed as a cause only by malignant compression of the bile duct. Many basolateral and canalicular transport proteins are downregulated in critically ill patients. This response most likely reflects sensitivity to inflammatory stimuli. Alterations in hepatocellular enzyme expression and activity, including those modulating phase I and II metabolism, may lead to profound changes in endobiotic and xenobiotic detoxification. Ductal cholestasis is another characteristic of hepatic dysfunction in the critically ill. This abnormality is most often seen in association with prolonged shock, in which impaired arterial perfusion may result in ischemic injury to the biliary system. Although hepatocellular impairment is most often fully reversible, ductular damage can lead to persistent alterations that may progress to secondary sclerosing cholangitis, an underappreciated long-term sequel of critical illness that carries a very poor prognosis.

The incidence of excretory impairment is underestimated by traditional “static” measures, such as serum transaminases or bilirubin. In contrast, “dynamic” tests, such as solute clearance, are much more sensitive. Hepatobiliary transport systems are essential for the uptake and excretion of various compounds, including bile acids and xenobiotics, and this partial function is best monitored by a functional test such as dye excretion.

Impact of Liver Dysfunction on Critical Care Pharmacology

Liver disease induces complex changes in the handling of drugs. These alterations are often unpredictable and may affect pharmacokinetics and pharmacodynamics. Therefore administration of medications to these patients must be carefully evaluated and strictly controlled, a mandate that is especially important when dosing drugs with a narrow therapeutic index. The additional risk of (hepatotoxic) side effects must be taken into account, and extreme caution is recommended when such drugs are used. Whenever possible, therapeutic monitoring should be performed, and dosage should be adjusted based on measured pharmacokinetic and pharmacodynamics properties. For some drugs, such as sedatives or analgesics, titrating dosage according to clinical effect may be sufficient. However, in the critically ill, it is essential that initial underdosing be avoided. This issue is especially germane with respect to antibiotic therapy in septic patients, where early and effective drug levels are essential for patient survival. In general, the following recommendations for dosing of drugs eliminated by hepatic metabolism and excretion should be observed in critically ill patients:

1.

In drugs with a high extraction ratio (>0.6), oral/enteral application leads to high first-pass metabolism and therefore low bioavailability. A reduction in hepatic blood flow due to shunting (intrahepatic, extrahepatic, or artificial following transjugular intrahepatic portosystemic shunt) or cirrhosis may substantially increase the bioavailability of such drugs. Thus, reducing initial doses and titrating maintenance doses should be considered for drugs with a high hepatic extraction ratio.
2.

Intravenous administration of drugs is usually more reliable and predictable. If liver blood flow is reduced, then maintenance doses should be reduced. Conversely, administration of prodrugs that have to be metabolized to their active form in the liver can result in reduced availability of the active drug (e.g., clopidogrel, enalapril).
3.

For drugs with a low hepatic extraction ratio (<0.3), clearance is dependent on the intrinsic capacity of the elimination pathway and on the fraction of drug that is not protein bound. Impairment of specific elimination mechanisms must be taken into account. For drugs with protein binding less than 90%, maintenance, but not initial, doses should be reduced. The amount of reduction can be estimated based on the severity of liver dysfunction. For example, in patients with liver dysfunction consistent with Child-Turcotte-Pugh (CTP) Class A, doses should be dropped to 50% of normal. A decrease to 25% of the normal dose is recommended in Class B dysfunction, whereas the more severe impairment associated with Class C requires the use of drug monitoring. These recommendations have limitations because the CTP score represents a rough approximation of impairment and each individual elimination mechanism may be differentially affected. Because phase II metabolism may be less severely impaired, drugs solely metabolized via this pathway should be preferentially used.
4.

For drugs with a low hepatic extraction ratio and high protein binding, changes in pharmacokinetics are unpredictable. Therefore drug monitoring is recommended wherever possible, and the unbound fraction should be measured.
5.

For hydrophilic substances, such as β-lactam antibiotics, an increased initial dosage should be considered in patients with ascites and edema because the volume of distribution may be substantially increased. However, for maintenance dosage of these drugs, the potential for renal dysfunction must be considered.

Recent reports suggest that liver function tests can be used to predict the required drug dosages. For example, the disappearance rate of indocyanine green from the plasma may be used to determine the appropriate argatroban maintenance dose in critically ill patients with heparin-induced thrombocytopenia type II, whereas the methacetin breath test can predict the increase in tacrolimus trough levels after liver transplantation. In the future, such testing might be used to govern therapy in which monitoring drug levels is not possible or where correct determination of the initial dose is crucial with respect to toxicity/side effects or therapeutic effect.

Only gold members can continue reading. Log In or Register to continue