1Nuffield Department of Anaesthetics, University of Oxford, Oxford, UK
The liver has a number of separate yet integrated functions. Assessment of hepatic behaviour during anaesthesia and the perioperative period generally involves tests that assess only part of the livers overall function, and assumes that behaviour in one area of activity reflects its function in other areas.
The functions of the liver can broadly be broken down into seven main areas: catabolic and anabolic functions with respect to carbohydrate, fat, and protein metabolism; production of bile; production and excretion of bilirubin; immunological functions involving the production and release of cytokines and interferons; scavenging and filtration of endotoxins and bacteria; storage of vitamin B12 and glycogen; and the biotransformation and elimination of drugs and xenobiotics.
While the hepatologist may use the prothrombin ratio or perhaps plasma pre-albumin concentrations (half-time 12 days) as a marker of the well-being of the liver for patients in acute liver failure, the chronic liver patient is best monitored by measurement of plasma enzymes (transaminases and gamma-glutamyl transpeptidase), serum bilirubin, and the plasma protein albumin (with its longer half-time of about 20 days). Traditionally, the anaesthetist has assessed the effects of drugs on the liver by measurement of the release into the blood or plasma of hepatic enzymes (such as the transaminases, alkaline phosphatase, and gamma-glutamyl transpeptidase, and more specific enzymes such as ornithine transcarbamoylase and 5'-nucleotidase).1 More recently, attention has been turned to pharmacological tests of liver functionbased on either the clearance of marker substances or measurement of the pharmacological effects of drugs that are wholly eliminated by the liver.
Assessment of liver function by drug disposition requires the ideal agent to display all or some the following characteristics.2
1. Non-toxic, and without any pharmacological effect.
2. Able to be administered i.v., or else rapidly and completely absorbed after oral dosing.
3. The rate-limiting step of metabolism should be affected by liver disease (i.e. avoid drugs that are metabolized at other sites apart from the liver).
4. Parent compound, its metabolites and/or both should be readily measurable in biological fluids, saliva, or breath.
5. The drug should not be highly protein bound in plasma or tissues unless there is an accompanying high hepatic extraction ratio.
6. Inexpensive!
7. Reliable assays available for drug and/or metabolites.
8. Agent available and licensed for human use.
Drugs used in disposition studies to assess liver function can be classified according to the rate-limiting step in their elimination, and hence we can subdivide the tests into those examining the effects of liver dysfunction on hepatic blood flow, intrinsic clearance, or unbound fraction.3
Capacity-limited, binding-insensitive hepatic elimination
These drugs have a low intrinsic clearance relative to total hepatic blood flow, and binding is usually less than 30% to plasma proteins. Hence, drug clearance will not be affected by changes in plasma protein binding or liver blood flow. The effect of liver disease on dysfunction of these drugs will be more straightforward than on the disposition of drugs with flow- or protein binding-limited clearance.
Examples of drugs in this group include phenazone, aminophenazone, theophylline, caffeine, and many of the neuromuscular blocking drugs (all of which are metabolized by hepatic oxidation). Those metabolized by conjugation include frusemide, morphine, oxazepam, and temazepam. Glucuronidation appears to be relatively unaffected by liver impairmentamong the possible reasons why this occurs are: activation of latent glucuronyl transferase enzyme(s); extra-hepatic glucuronidation in liver disease;4 presence of large reserves of intra- and extra-hepatic glucuronyl transferase; location of the glucuronidation enzymes in parts of the liver lobule less affected by liver disease; or a reduction in enterohepatic recycling of glucuronides which would tend to increase the clearance of the parent drug and thereby offset the decreased glucuronide formation.5
There are fewer data on the effects of liver disease on other conjugation pathways, although studies show impairment of acetylation in liver disease (e.g. isoniazid,6 procainamide,7 and sulfadimidine8 9), and acetaminophen sulfation being reduced in cirrhosis.10
Another drug whose disposition is altered in liver disease is antipyrine; it has a reduced clearance in cirrhosis and viral hepatitis, and in obstructive jaundice the antipyrine half-life is a valuable predictor of outcome.11 Aminopyrine (assessed as the 14CO2 aminopyrine breath test) is also negligibly protein bound, and like antipyrine assesses hepatic metabolizing capacity rather than liver blood flow. Although it has been used in the pig to assess the outcome of transplantation,12 the aminopyrine test has a number of limitations to its use in man. For instance, reduction in the test output can occur in patients with sepsis and viral infections; there is administration of radioactivity; endogenous carbon dioxide production may interfere with the results; and the drug can induce agranulocytosis in patients sensitized previously by pyrazolone derivatives.
Capacity-limited, but binding-sensitive hepatic elimination (for example digitoxin, mexilitine, midazolam, erythromycin, and tolbutamide)
Usually these drugs show low intrinsic clearances, but high plasma protein binding (>85%), although there are some drugs with a large hepatic intrinsic clearance despite very high binding to plasma proteins.13 It is difficult to generalize on the effects of liver dysfunction on the elimination of drugs in this group.
Flow-limited elimination
In these cases, hepatic clearance is high relative to liver blood flow (>70%) and will be influenced by the latter. For drugs that are highly bound to albumin or other proteins, elimination is not limited to the unbound drug alone (i.e. clearance is unrestrictive). Examples of drugs with these characteristics include ICG, galactose, D-sorbitol, propranolol, lidocaine, pethidine (meperidine), and morphine. It is members of this latter group that have been most widely used to date for assessing graft function in the liver transplant patient, although with variable reliability and sensitivity.1419
The determination of a drugs pharmacokinetic profile can provide much information about the metabolic capacity of the liver. However, these tests do not always provide the clinician with the aetiology of the observed reduction in drug clearance (the product of flow and extraction ratio).
Measurement of plasma drug or metabolite concentrations may not be easily achieved, and certainly not on-line, so that the attention of clinical pharmacologists to the measurement of dynamic endpoints has been encouraged. However, for the opioid and hypnotic drugs, we have no easily measurable dynamic endpoint, although researchers have sought surrogate markers such as the median power frequency or spectral edge frequency of the EEG and, more recently, the BIS monitor.
Other disadvantages of some of these drugs include: the potential for interference in the elimination pathways by other drugs metabolized by the same CYP 450 (e.g. lidocaine); clearance affected by both changes in cellular function and flow (e.g. ICG); caffeine elimination is affected by smoking; extra-hepatic elimination pathways exist for galactose; antipyrine clearance is affected by various environmental and genetic factors, and other drugs; and the aminopyrine and erythromycin tests contain the need for the potential hazard of radioactivity.
Another approach has been to use neuromuscular blocking drugs, which have a number of distinct advantages over other markers as they are routinely used during anaesthesia and surgery. The liver is predominantly the route of elimination of aminosteroidal neruomuscular blocking drugs, and there is a close relationship between blood drug concentration and the degree of neuromuscular paralysis in both healthy patients and those with hepatic disease.2022 Furthermore, measurement of the dynamic endpoint of relaxation (by single twitch or the train-of-four twitch response) is much easier.
Early perioperative assessment of transplant graft function by examining the disposition and dynamics of neuromuscular blocking drugs was first studied by Lukin and colleagues,23 and Marcel and colleagues24 using the aminosteroids vecuronium and rocuronium. Both groups of authors found a significant correlation between recovery from neuromuscular block and graft function in the early postoperative period. Prolongation of the recovery time was associated with primary graft dysfunction. One advantage of the use of neuromuscular block as a kinetic marker is the relative absence of hysteresis between drug concentration and effect.25 Thus, in the absence of active metabolites, which rocuroniumunlike vecruoniumdoes not possess, the kinetics of the neuromuscular blocking drugs can be followed by the train-of-four twitch response.
What are the main messages of the paper by Gao and colleagues26 in this issue of the Journal? There are changes in drug disposition during the anhepatic and neohepatic phases of orthotopic transplantation, which seem to correlate with early postoperative graft function. The reduction postoperatively in the plasma relaxant concentration is most likely the result of relaxant excretion by the newly perfused liver. In one respect, their data, therefore, tally with those of Marcel and colleagues24 who found a close relationship between recovery neuromuscular block and early postoperative graft function. The present authors suggest that use of rocuronium as a pharmacological probe during liver transplantation by measurement of neuromuscular block together with infusion dose requirements,26 may offer a useful approach towards telling the clinician about the likely immediate function of the grafted organwe await further confirmatory studies with interest!
However, the dynamics of the neuromuscular blocking drugs may be affected by a number of disturbances in the perioperative period such as: acidbase status upsets (acidosis prolongs neuromuscular block); inhaled anaesthetics also prolong neuromuscular block, and may affect drug disposition; electrolyte imbalances such as decreases in Ca2+ or K+ enhance neuromuscular block; while mild hypothermia also significantly prolongs both vecuronium and atracurium block in man. These abnormalities, which will affect drug dynamics more than kinetics, may increase the inter-individual variability that occurs and so prevent a clear separation of those patients with and those without good postoperative hepatic function.
References
1 Sear JW, Prys-Roberts C, Dye A. Hepatic function after anaesthesia for major vascular reconstructive surgery: a comparison of four anaesthetic techniques. Br J Anaesth 1983; 55: 6037[Abstract]
2 Gao L, Ramzan I, Baker AB. Potential use of pharmacological markers to quantitatively assess liver function during liver transplantation surgery. Anaesth Intens Care 2000; 28: 37585[ISI][Medline]
3 Blaschke TF. Protein binding and kinetics of drugs in liver disease. Clin Pharmacokinet 1977; 2: 3244[ISI][Medline]
4
Hoyumpa AH, Brown MW, Johnson R, Troxell R, Snowdy P, Schenker S. Effects of liver disease on the disposition of ciramadol in humans. J Clin Pharmacol 1989; 29: 21724
5 Ghabrial H, Desmond PV, Watson KJR, et al. The effects of age and chronic liver disease on the elimination of temazepam. Eur J Clin Pharmacol 1986; 30: 937[ISI][Medline]
6 Acocella G, Bonollo I, Garimoldi M, Mainairdi M, Tenconi IT, Nicolis FB. Kinetics of rifampicin and isoniazid administered alone and in combination to normal subjects and patients with liver disease. Gut 1972; 13: 4753[ISI][Medline]
7 Du Souich P, Erill S. Metabolism of procainamide and p-amino-benzoic acid in patients with chronic liver disease. Clin Pharmacol Ther 1977; 22: 58895[ISI][Medline]
8 Horvath T, Par A, Past T, Tapsonyi Z, Kadas I. Disorders of biotransformation during progression of alcoholic liver disease. Acta Medica Hungarica 1986; 43: 3517[Medline]
9 Horvath T, Past T, Par A, Kadas SI, Javor T. Sulfadimidine kinetics in patients with various chronic liver diseases. Polish J Pharmacol Pharmacy 1988; 40: 25764
10 Forrest JAH, Finlayson NDC, Adjepon-Yamoah KK, Prescott LF. Antipyrine, paracetamol and lignocaine elimination in chronic liver disease. BMJ 1979; 1: 13847
11 McPherson GAD, Benjamin IS, Boobis AR, Blumgart LH. Antipyrine elimination in patients with obstructive jaundice: a predictor of outcome. Am J Surg 1985; 149: 1403[ISI][Medline]
12 Oakland CDH, Hickman R, Terblanche J. The aminopyrine breath test predicts the outcome of hepatic transplantation in pigs. Hepatology 1989; 9: 6025[ISI][Medline]
13 Byrne AJ, Morgan DJ, Harrison PM, McLean AJ. Variation in hepatic extraction rate with unbound drug fraction: discrimination between models of hepatic drug elimination. J Pharmaceutical Sci 1985; 74: 2057[ISI][Medline]
14 Olinga P, Maring JK, Groothuis GMM, et al. Value of the in vitro or in vivo mono-ethylglycinexylidide test for predicting liver graft function. Transplantation 1997; 64: 605[ISI][Medline]
15 Dette K, Knoop M, Langrehr JM. Donor MEGX test fails to predict graft liver function after orthotopic liver transplant. Transplantation Proc 1997; 29: 3767[ISI][Medline]
16 Freys G, Pottecher T, Calon B, et al. Early assessment of transplanted liver function: lignocaine clearance test (MEGX). Eur J Anaesthesiol 1997; 14: 397405[ISI][Medline]
17 Jalan R, Plevris JN, Jalan AR, Finlayson NDC, Hayes PC. A pilot study of ICG clearance as an early predictor of graft function. Transplantation 1994; 58: 196200[ISI][Medline]
18 Nagel RA, Dirix LY, Hayllar KM, Preisig R, Tredger JM, Williams R. Use of qualitative liver function testcaffeine clearance or galactose elimination capacityafter orthotopic liver transplantation. J Hepatol 1990; 10: 14957[ISI][Medline]
19 Lamesch P, Ringe R, Oellerich M, et al. Assessment of liver function in the early postoperative period after liver transplantation with ICG, MEGX and galactose tests. Transplantation Proc 1990; 22: 153941[ISI][Medline]
20 Agoston S, Crul JF, Kersten UW, Scaf AH. Relationship of the serum concentration of pancuronium to its neuromuscular activity in man. Anesthesiology 1977; 47: 50912[ISI][Medline]
21 Wierda JM, Proost JH, Schiere S, Hommes FD. Pharmacokinetics and pharmacokinetic/dynamic relationship of rocuronium bromide in humans. Eur J Anaesthesiol 1994; 11 (Suppl. 9); 6674
22 Khalil M, DHonneur G, Duvaldestin P, Slavov V, De Hys C, Gomeni R. Pharmacokinetics and pharmacodynamics of rocuronium in patients with cirrhosis. Anesthesiology 1994; 80: 12417[ISI][Medline]
23 Lukin CL, Hein HAT, Swygert TH, et al. Duration of vecuronium-induced neuromuscular blockade as a predictor of liver allograft dysfunction. Anesth Analg 1995; 80: 52633[Abstract]
24 Marcel RJ, Ramsay MAE, Hein HAT, et al. Duration of rocuronium-induced neuromuscular block during liver transplantation: a predictor of primary allograft function. Anesth Analg 1997; 84: 8704[Abstract]
25 Parker CJ, Hunter JM. Dependence of the neuromuscular blocking effect of atracurium upon its disposition. Br J Anaesth 1992; 68: 55561[Abstract]
26
Gao L, Ramzan T, Baker B. Rocuronium plasma concentrations during three phases of liver transplantation: relationship with early postoperative graft liver dysfunction. Br J Anaesth 2002; 88: 76470