Analgesia in Liver Impair Pts

8/18/2019 Analgesia in Liver Impair Pts

1/25

Analgesics in Patients with HepaticImpairmentPharmacology and Clinical Implications

Marija Bosilkovska,1 Bernhard Walder,2 Marie Besson,1 Youssef Daali1 and Jules Desmeules1

1 Division of Clinical Pharmacology and Toxicology, Geneva University Hospitals, Geneva, Switzerland

2 Division of Anesthesiology, Geneva University Hospitals, Geneva, Switzerland

Contents

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16461. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16462. Hepatic Dysfunction and Drug Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1647

2.1 Hepatic Clearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16472.2 Pharmacokinetic Changes in Liver Disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1648

3. Analgesics in Patients with Hepatic Impairment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16493.1 Paracetamol (Acetaminophen) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1649

3.1.1 Hepatotoxicity and Safety Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16493.1.2 Pharmacokinetic Changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16523.1.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1652

3.2 Nonsteroidal Anti-Inflammatory Drugs (NSAIDs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16523.2.1 Pharmacodynamic Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16523.2.2 NSAID-Induced Liver Injury. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16533.2.3 Pharmacokinetics of Specific NSAIDs in Hepatic Impairment . . . . . . . . . . . . . . . . . . . . . 1654

3.3 Opioids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16553.3.1 Codeine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16553.3.2 Tramadol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16563.3.3 Tapentadol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16563.3.4 Oxycodone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16573.3.5 Morphine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16573.3.6 Hydromorphone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1658

3.3.7 Pethidine (Meperidine) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16583.3.8 Methadone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16593.3.9 Buprenorphine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1659

3.3.10 Fentanyl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16603.3.11 Sufentanil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16603.3.12 Alfentanil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16613.3.13 Remifentanil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661

3.4 Neuropathic Pain Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16623.4.1 Antidepressants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16623.4.2 Anticonvulsants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16623.4.3 Opioids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663

4. Conclusion and Clinical Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1663

REVIEW ARTICLE Drugs 2012; 72 (12): 1645-1669

0012-6667/12/0012-1645/$55.55/0

Adis ª 2012 Springer International Publishing AG. All rights reserved.


2/25

Abstract The physiological changes that accompany hepatic impairment alter drugdisposition. Porto-systemic shunting might decrease the first-pass metabo-

lism of a drug and lead to increased oral bioavailability of highly extracteddrugs. Distribution can also be altered as a result of impaired production of drug-binding proteins or changes in body composition. Furthermore, theactivity and capacity of hepatic drug metabolizing enzymes might be affectedto various degrees in patients with chronic liver disease. These changes wouldresult in increased concentrations and reduced plasma clearance of drugs,which is often difficult to predict.

The pharmacology of analgesics is also altered in liver disease. Pain manage-ment in hepatically impaired patients is challenging owing to a lack of evidence-based guidelines for the use of analgesics in this population. Complications suchas bleeding due to antiplatelet activity, gastrointestinal irritation, and renal

failure are more likely to occur with nonsteroidal anti-inflammatory drugs inpatients with severe hepatic impairment. Thus, this analgesic class should be

avoided in this population.The pharmacokinetic parameters of paracetamol (acetaminophen) are

altered in patients with severe liver disease, but the short-term use of this drugat reduced doses (2 grams daily) appears to be safe in patients with non-alcoholic liver disease.

The disposition of a large number of opioid drugs is affected in the pre-sence of hepatic impairment. Certain opioids such as codeine or tramadol, forinstance, rely on hepatic biotransformation to active metabolites. A possiblereduction of their analgesic effect would be the expected pharmacodynamic

consequence of hepatic impairment. Some opioids, such as pethidine (me-peridine), have toxic metabolites. The slower elimination of these metabolitescan result in an increased risk of toxicity in patients with liver disease, andthese drugs should be avoided in this population.

The drug clearance of a number of opioids, such as morphine, oxycodone,tramadol and alfentanil, might be decreased in moderate or severe hepaticimpairment. For the highly excreted morphine, hydromorphone and oxycodone,an important increase in bioavailability occurs after oral administration in

patients with hepatic impairment. Lower doses and/or longer administrationintervals should be used when these opioids are administered to patients withliver disease to avoid the risk of accumulation and the potential increase of

adverse effects. Finally, the pharmacokinetics of phenylpiperidine opioidssuch as fentanyl, sufentanil and remifentanil appear to be unaffected in he-patic disease. All opioid drugs can precipitate or aggravate hepatic en-cephalopathy in patients with severe liver disease, thus requiring cautious useand careful monitoring.

1. Introduction

The liver has a predominant role in the pharm-acokinetics of most drugs. Therefore, drug dis-

position may be altered in patients with hepaticimpairment. Liver dysfunction is often progressive,and drug elimination impairment increases along

with the increase in liver dysfunction. In patientswith certain types of hepatic dysfunction, such aschronic active hepatitis or liver cancer withoutcirrhosis, drug elimination is altered only to a

small extent.[1,2]Unlike estimates of glomerular filtration rate

(GFR; creatinine or inulin clearance), which are

1646 Bosilkovska et al.

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (12)


3/25

useful for determining the pharmacokinetic pa-rameters of drug elimination in renal impairment,no adequate biomarkers relating to hepatic func-tion and drug elimination capacity are available.Various classification schemes and dynamic liverfunction tests have been developed to predict drughandling in patients with liver disease. The mostcommonly used systems to scale the severity of hepatic impairment are the Child-Pugh classifi-cation and the Model for End-Stage Liver Dis-ease (MELD) system.[3] The Child-Pugh systemincorporates three measurable laboratory vari-

ables (serum bilirubin, albumin and prothrombintime) and two clinical variables (the presence of ascites and encephalopathy). Disease severity isclassified as mild, moderate or severe (Child-Pughclasses A, B and C, respectively). The MELD sys-tem is based on serum bilirubin and creatinineconcentrations, the international normalized ratioof prothrombin time, and the underlying cause of liver disease.[3]

The US Food and Drug Administration andthe European Medicines Agency have issued

directives encouraging industries to conductpharmacokinetic studies in patients with hepaticimpairment for drugs likely to be used in thesepatients or for drugs for which hepatic impairmentmight significantly affect pharmacokinetics.[4,5]

Despite these directives, dosage adjustment recom-mendations in patients with hepatic impairment areoften lacking for older drugs, which is the case formost of the commonly used analgesics.

Pain relief is central to improving the qualityof life of every patient, including patients with

liver disease. Thus, analgesics are likely to be usedfrequently in patients with hepatic impairment.The metabolism and elimination of the majority of analgesics, including paracetamol (acetaminophen),nonsteroidal anti-inflammatory drugs (NSAIDs)and opioids, can be impaired in patients with liverdisease. Drug accumulation and increased sideeffects might occur as a consequence of this im-pairment. In addition to modifying pharmaco-kinetics, liver disease can also substantially alterpharmacodynamic effects. An example detailed

later in this review is increased sensitivity to opi-oids, which can cause cerebral dysfunction oraggravate pre-existing hepatic encephalopathy.[6]

Liver disease alters pharmacokinetics, but drugsthemselves can also impair liver function. Drug-induced liver injury is a potential complication of most drugs, including analgesics.[7] Acute liverfailure is a well known adverse effect of high-doses of paracetamol, one of the most widely usedanalgesics.[8] Hepatotoxicity has also been de-scribed in patients using acetylsalicylic acid orNSAIDs.[9] Some drugs in this class, such as ni-mesulide or diclofenac, are more likely to provokehepatic injury than others.[10,11] The new moreselective cyclooxygenase (COX)-2 inhibitors, such

as lumiracoxib, were also associated with hepato-toxicity.[12] Although rare, NSAID-induced liverinjury should not be underestimated owing to thecommon and widespread use of these drugs in thepopulation.

A paucity of evidence exists regarding the safetyand efficacy of pharmacological pain therapiesin patients with hepatic impairment.[13] Physiciansdisplay significant variability in their recommend-ations for the use of analgesics in this population.Healthcare providers often consider the use of an-

algesics in patients with cirrhosis as unsafe, leadingto under-treatment of pain in this population.[14]

The aim of this review is to resume and analyse thepublished data on various analgesics in patientswith hepatic impairment and provide evidence forthe safe use of these drugs in this population.

2. Hepatic Dysfunction andDrug Pharmacokinetics

2.1 Hepatic Clearance

Hepatic metabolism is the main eliminationpathway for most lipophilic drugs. The efficiencyof drug removal by the liver, so-called hepaticclearance, is determined by hepatic blood flow,plasma protein binding and intrinsic clearance,which represent the metabolic activity of hepaticenzymes. Hepatic clearance may be describedwith equation 1:

CLH ¼QH E H (Eq: 1Þ

where QH is the hepatic blood flow and EH is thehepatic extraction ratio, which depends on liverblood flow (QH), intrinsic clearance (CLint) and

Analgesics in Hepatic Impairment 1647



4/25

unbound drug fraction (f u). Thus, equation 1 canbe presented as equation 2:

CLH ¼QH f uCLint

QH þ f uCLint ðEq: 2Þ

The hepatic clearance of drugs with high ex-traction ratios (EH > 0.7) depends largely on liverblood flow. A decrease in liver blood flow orthe presence of intra- and extrahepatic porto-systemic shunting may strongly affect the clear-ance of these drugs. In contrast to that of highlyextracted drugs, the hepatic clearance of poorly

extracted drugs (EH < 0.3) is mainly influenced bychanges in the plasma protein binding and in-trinsic metabolic clearance, as shown in equation2. The effect of liver disease on these parametersis discussed below. Hepatic clearance for drugswith an intermediate extraction ratio may be af-fected by liver blood flow, plasma protein bind-ing and metabolic activity.[3,15]

2.2 Pharmacokinetic Changes in Liver Disease

Orally administered drugs absorbed from thegastrointestinal tract pass into the portal vein andcan undergo substantial metabolism in the liverbefore reaching the systemic circulation, a phe-nomenon known as the first-pass effect. Cirrhosismay lead to porto-systemic shunts and develop-ment of collateral circulation. A substantial frac-tion of the blood, which should normally reachthe portal vein, may flow through this collateralcirculation, reducing mesenteric blood flow throughthe liver. Drugs with intermediate or high hepatic

extraction ratios have increased oral bioavailabilityin patients with cirrhosis as a result of reduced first-pass metabolism.[16,17] Increased bioavailabilitycombined with the decreased hepatic clearancediscussed below can cause an important increasein the area under the plasma concentration-timecurve (AUC).[18]

One characteristic of liver disease, especiallycirrhosis, is impaired production of drug-bindingproteins such as albumin and a1-acid glycopro-tein. Decreased levels of these proteins cause an

increase in the free fraction of drugs, which isparticularly important for highly bound drugs(f u< 0.1). Because only the unbound fraction of a

drug can enter or leave tissue compartments, de-creases in plasma protein binding influence drugdistribution, increasing the distribution volume(Vd) of certain drugs.

[3,15] The difference betweentotal plasma clearance and plasma clearance of the unbound fraction is crucial in interpretationsof pharmacokinetic data for highly bound drugsin patients with liver disease. In some cases, totaldrug clearance may appear to be unimpaired inthese patients even though clearance of the un-bound fraction is markedly reduced. In fact, thedecrease in metabolic capacity present in liver

disease is counterbalanced by the increase of freedrug fraction, leading to the false conclusion thatdrug metabolism is unaffected. The values fortotal drug plasma concentrations and clearanceare normal, but the clearance of the unboundfraction is reduced because more of the free drugenters the tissues (increased distribution).[3,19,20]

With the progression of liver disease, changes inbody composition such as increased extracellularfluid (ascites, oedema) and decreased musclemass occur, altering the Vd.

[6]

The hepatic metabolism of drugs is dividedinto two types and steps of biotransformations:phase I and phase II. Phase I reactions are oxido-reductive processes mainly catalysed by monooxy-genases such as cytochrome P450 (CYP) enzymes,whereas phase II reactions are catalysed by con-

jugating enzymes. The function and expressionof these enzymes can be altered in patients withliver disease. Phase I enzymes are generally con-sidered to be more affected in liver disease thanare phase II enzymes, likely owing to the higher

sensitivity of phase I enzymes to hypoxia causedby shunting, sinusoidal capillarisation and reducedperfusion.[21,22] Isoforms of CYP are affected to vari-able degrees depending on the severity of liver dis-ease. Frye et al.[23] have found a strong decrease inthe metabolic activity of CYP2C19 in patients withmild liver disease, whereas CYP1A2, CYP2D6and CYP2E1 activity in these patients seemedrelatively preserved. However, patients with mod-erate to severe liver disease displayed decreasedmetabolic activity for all of the CYP isoforms

studied. The type of liver disease (cholestatic, hepa-tocellular or metastatic) also influences the degreeof CYP metabolic activity impairment.[19]




5/25

As previously mentioned, phase II reactions,especially glucuronidation, are affected by liverimpairment to a lesser extent. Possible explana-tions for this difference may be the up-regulationof uridine 50-diphosphate glucuronosyltransferase(UGT) activity in the remaining hepatocytes,[24] afavourable localization of the glucuronyltransferasesin the microsomes,[20] or increased extrahepaticmetabolism.[25] With some drugs, however, glu-curoconjugation can be preserved in the presenceof mild or moderate liver disease but altered inpatients with severe disease.[20] The biliary clear-

ance of some drugs or metabolites eliminated bybiliary excretion can be reduced in patients withliver disease, requiring dose reduction or avoid-ance of these drugs. However, studies of this effectare rather limited.[1,3]

Renal function often becomes impaired in pa-tients with severe liver disease. The renal impair-ment that occurs in severe liver disease withoutany laboratory, anatomical or clinical evidence of another cause is called hepatorenal syndrome.[26]

Patients with this syndrome can display sig-

nificantly diminished renal drug clearance. Im-paired renal function and drug clearance can alsooccur in patients with mild to moderate liver dis-ease and is often underestimated because serumcreatinine levels in these patients do not riseeven when the GFR is very low.[27] This pheno-menon might occur due to the underproductionof creatinine when muscular mass is diminishedor as a result of decreased hepatic production of creatine, the substrate for creatinine produc-tion.[28] Besides the serum creatinine level, both

the measured and the calculated creatinine clear-ance (using the Cockcroft and Gault method[29])predict GFR adequately in cirrhotic patients withnormal renal function but overestimate GFR incirrhotic patients with renal impairment.[30] Thisinformation must be considered when assessingrenal function and prescribing drugs with pre-dominantly renal elimination in hepatically im-paired patients.

The pharmacokinetic changes described aboveare mostly observed in cirrhotic patients. In patients

with chronic liver disease, but without significantfibrosis, drug pharmacokinetics are unchanged ormodified only to a small extent.[2]

3. Analgesics in Patients with HepaticImpairment

Pain management in hepatically impaired pa-tients is challenging because evidence-based guide-lines for the use of analgesics in this population arelacking. Table I summarizes the findings on thedisposition of analgesics in hepatic impairment andprovides practical recommendations on the use of these drugs in this population.

3.1 Paracetamol (Acetaminophen)

3.1.1 Hepatotoxicity and Safety Issues

Paracetamol is commonly recommended as afirst-choice analgesic for various nociceptiveacute or chronic pain conditions and remains oneof the safest accessible analgesics for multimorbidpatients. However, the use of paracetamol in pa-tients with hepatic disease is often avoided, prob-ably owing to the well known association betweenparacetamol overdose and hepatotoxicity. Para-cetamol is mainly metabolized to glucuronide andsulphate conjugates, and a small proportion (


6/25

remains unclear whether therapeutic doses of paracetamol would be more toxic in patients withchronic liver disease or cirrhosis.[66,68]

A double-blind, two-period, crossover studywas conducted in 20 patients with chronic liverdisease to analyse the development of adverse

Table I. Pharmacokinetic alterations and recommendations for the use of analgesics in hepatic impairment

Analgesic Pharmacokinetics changes in patients with liverdisease Recommendations and dose adjustments

a

Paracetamol

(acetaminophen)b50 – 100% › t½; › AUC; fl CL

[31-34] Reduce doses to 2 g/daily

Nonsteroidal anti-inflammatory drugsc

Aspirin

(acetylsalicylic acid)2-fold › AUC of salicylic acid fu; higher risk of

salicylate toxicity[35]

Naproxen fl CLU by 60%[36] Reduce doses by 50%

Ibuprofen No significant changes[37] No adjustment

Etodolac No significant changes[38] No adjustment

Sulindac 3-fold › AUC for sulindac and 4-fold › AUC for

sulindac sulfide (active metabolite)[37]Reduce doses

Diclofenac No changes or possible › AUC in alcoholic cirrhosis No adjustment

Celecoxib 40% › AUC in mild and 140% › AUC in moderate

liver disease[39]Moderate liver disease: reduce doses by 50%

Opioids

Codeine Reduc ed transformation to morphine Avoid use, possible lack of analgesic effects

Tramadol 3.2-fold › AUC and 2.6-fold › t½, lack of

transformation to O-demethyl tramadol[40]Prolong dosage intervals or reduce doses. Analgesic effects

not evaluated in this population

Tapentadolb 1.7- and 4.2- fold › AUC and 1.2- and 1.4-fold › t½ in

mild and moderate liver disease, respectively[41]Moderate liver disease: low doses and prolonged dosing

interval

Severe liver disease: no data available

Morphineb › Oral bioavailability; › t½; fl CL[42-45] 2-fold prolongation in dosage intervals. If administered orally

also reduce doses

Oxycodone › AUC; › t½; fl CL[46,47] Use lower doses with prolonged dosage intervals

Hydromorphoneb › Oral bioavailability; no changes in t½ in moderate liver

disease[48]Reduce doses, consider dosage interval prolongation only in

severe liver disease

Pethidine › oral bioavailability; 2-fold › t½; 2-fold fl CL[49-52] Avoid repeated use, risk of neurotoxic metabolite

accumulation

Methadone › t½; possible risk of accumulation[53,54] No changes needed in mild and moderate liver disease

Careful titration in severe liver disease

Buprenorphine No data. Possible fl of its metabolism No recommendations

Fentanyl No changes after single IV dose in moderate

liver disease[55]Dose adjustment usually not needed, might be necessary if

continuous infusion or transdermal patches are used

Sufentanil No changes after single IV dose in moderate

liver disease[56]

Dose adjustment usually not needed, might be necessary in

continuous infusionAlfentanil fl Protein binding; › t½; fl CL even in patients with mild

liver disease[57-59]Reduce dose and prolong dosing interval

Prefer another phenylpiperidine opioid

Remifentanil No changes[60,61] No adjustment

a Refers to dose adjustment in severe liver disease unless indicated otherwise.

b Analgesics metabolized by conjugation.

c Dose adjustments refer to patients with mild to moderate liver disease. In patients with severe liver disease NSAIDs should be avoided due

to their antiplatelet activity, gastrointestinal irritation and increased renal toxicity.

AUC= area under the plasma concentration-time curve; CL= total plasma clearance; CLU= clearance of the unbound drug fraction;

fu= unbound drug fraction; IV = intravenous; t½= elimination half-life; ›› indicates increase; flfl indicates decrease.




7/25

reactions and the deterioration of liver-relatedlaboratory tests (e.g. levels of bilirubin, alkalinephosphatase, serum bile acids, creatinine, albuminand prothrombin time). Patients were randomlyassigned to either paracetamol 4 g or placebo dailyfor 13 days, after which they were crossed over tothe alternate treatment for 13 days. Comparedwith placebo, the use of paracetamol during thisperiod appeared to have no significant effect onclinical features or laboratory tests.[69] The find-

ings of a case-control study evaluating the im-plication of over-the-counter analgesics in acutedecompensation in patients with cirrhosis sug-

gested no association between the occasionaluse of low-dose paracetamol (2 – 3 g/day) and thedecompensation of cirrhosis.[70] The number of studies evaluating the safety of paracetamol inpatients with liver disease without cirrhosis israther limited. In a randomized controlled trial,Dargere et al.[71] found no difference in the varia-tion of serum levels of alanine transaminase (ALT)between patients with non-cirrhotic chronic hepa-titis C receiving paracetamol 3 g/day or placebo for7 days.

An especially delicate and often controversial

question is the use and hepatotoxicity of ther-apeutic doses of paracetamol in chronic alcoholusers. Glutathione levels are known to be reduced inchronic alcohol consumers or fasting subjects.[72,73]

It is also known that the CYP2E1 isoenzyme re-sponsible for the metabolism of paracetamol tothe toxic intermediate NAPQI is induced bychronic alcohol consumption.[74] It is thereforenot surprising that the production of NAPQI,estimated from the urinary concentration of cys-teine and N-acetylcysteine conjugates, is higher

in non-cirrhotic chronic alcohol users than insubjects who do not consume alcohol.[75] Thismakes chronic alcohol users (cirrhotic or not)more vulnerable to elevated doses of paraceta-mol. Many reports, mostly retrospective studiesor case reports, have found an association betweenalcohol use and enhanced paracetamol toxicity incases of overdose but also when paracetamol wasused at therapeutic doses.[76-79] A randomizedplacebo-controlled study demonstrated no increasein serum aminotransferases or international normal-

ized ratio in alcoholic subjects receiving therapeuticdoses of paracetamol (4 g/day) for 48 hours.[80]

Nevertheless, a more recent randomized placebo-controlled study has shown a small but signif-icant increase in ALT at the end of treatmentin moderate alcohol consumers taking paraceta-mol 4 g daily for 10 days. Serum ALT levels in-creased from 21.3 – 7.6 IU/L before treatment to30.0 – 19.6 IU/L at the end of the 10-day treat-ment period.[81] Although the clinical implica-tions of this elevation are unclear, precautions

should be taken if paracetamol is used in alco-holic patients, especially long term. The US Foodand Drug Administration requires a warning

NHCOCH3

O-Glucuronide

NHCOCH3

O-Sulfate

NHCOCH3

OH

Acetaminophen

NAPQI

Cysteine and N-acetyl cysteine conjugates

UGT

60%

NHCOCH3

OH

S-Glutathione

O

NCOCH3

SULT

30%

CYP2E1

GST

5%

Fig. 1. Simplifiedpresentationof paracetamol (acetaminophen) majormetabolic pathways. CYP= cytochrome P450; GST= glutathioneS-transferase; NAPQI =N-acetyl-p-benzoquinone imine, toxic inter-mediate metabolite; SULT = sulfotransferase; UGT=uridine 50-di-phosphate glucuronosyltransferase.




8/25

label for paracetamol-containing products stat-ing that individuals who consume three or morealcoholic beverages per day should consult theirphysician before using paracetamol.

3.1.2 Pharmacokinetic Changes

Pharmacokinetic studies in patients with livercirrhosis have shown an increase in the elimina-tion half-life (t½) of paracetamol ranging from50% to 100% compared with that in control sub-

jects. The AUC was significantly higher and plas-ma clearance of the drug was reduced, whereas the

mean values for maximum plasma drug concen-tration (Cmax) and the time to Cmax (tmax) did notdiffer.[31-34] In two of these studies, a correlationwas found between the t½ of the drug

[34] or plasmaclearance[31] and prothrombin time as well as theplasma albumin levels. In one of these studies, thet½ was doubled in patients with both low albu-min levels (1.4).[34] The other study showed that a10% decrease in prothrombin level decreased plas-ma clearance by 10%.[31] The correlation with al-

bumin levels was statistically less important.None of the studies showed correlation betweendrug t½ or plasma clearance and plasma bilirubinlevels.

The possible accumulation of repeatedly ad-ministered paracetamol in subjects with chronicliver disease has been evaluated in two stud-ies.[31,69] In both studies, six subjects receivedparacetamol 1 g four times per day over 5 days.No progressive accumulation of paracetamol wasapparent in the plasma of cirrhotic patients, de-

spite a slight prolongation of its t½.The production of the reactive hepatotoxic

intermediate NAPQI, estimated from the urinaryconcentration of cysteine and N-acetylcysteineconjugates, is enhanced in alcoholic subjects with-out cirrhosis but unaffected in cirrhotic subjectsabstaining from alcohol.[75] Another study con-firmed that the metabolic pattern in blood andurinary excretion did not differ between cirrhoticand healthy subjects after administration of asingle dose of paracetamol 1 g.[82]

In a study evaluating the pharmacokinetics of paracetamol in children with non-alcoholic fattyliver disease, higher concentrations of paracetamol

glucuronide were observed, although they did notseem to affect the rate of paracetamol eliminationbecause no difference in the pharmacokineticparameters for paracetamol itself was observedbetween children with non-alcoholic fatty liverdisease and healthy children. The cysteine andmercapturic acid conjugate concentrations werenot determined; therefore, it is difficult to evalu-ate whether the use of paracetamol in this pop-ulation increases the risk of hepatic injury.[83]

The pharmacokinetics of paracetamol in patientswith acute viral hepatitis without cirrhosis were

not significantly altered compared with that incontrol subjects. However, the t½ and the AUCincreased, and the plasma clearance decreased, insubjects during the acute hepatitis phase com-pared with that during convalescence. The au-thors suggest that patients with hepatitis can takeconventional doses of paracetamol, and prolongeddosage intervals are necessary only in serious casesin which prothrombin time is prolonged.[84]

3.1.3 Summary

In summary, the few available studies suggestthat the use of short-term therapeutic doses of paracetamol in patients with non-alcoholic cir-rhotic liver disease cause no accumulation ordeterioration of liver-related laboratory tests, in-dicating that this drug can be used in thesepatients at normal doses. However, owing to thechanges in the pharmacokinetics and the vulner-ability of this population, it seems reasonable tolimit the adult daily dose to 2 g, half the suggestedtherapeutic dose. Physicians should remain atten-

tive to any symptoms indicating a possible ag-gravation of the hepatic function. Doses shouldbe reduced to 2 g/day, or paracetamol should beavoided as much as possible in chronic alcoholusers.

3.2 Nonsteroidal Anti-Inflammatory Drugs(NSAIDs)

3.2.1 Pharmacodynamic Complications

Patients with severe liver disease, especiallythose with cirrhosis and ascites, display unstable

renal haemodynamics. Even with normal GFRand renal blood flow, renal perfusion in thesepatients is sensitive to modest reductions in plas-




9/25

ma volume. The impairment of renal functionand the effects of vasoconstrictive hormones arecountered by the increased production of vaso-dilatory renal prostaglandins in these patients.[85]

NSAIDs inhibit the compensatory actions of prostaglandins by inhibiting their synthesis. Thisinhibition decreases GFR and renal blood flow.These agents also reduce the capability of thekidneys to excrete sodium and water and can thusbe responsible for the formation of ascites.[86-88]

Renal impairment with reduced GFR, renal bloodflow, and sodium and water excretion occurs in

patients with liver disease when ibuprofen,

[89,90]

indomethacin,[91] aspirin (acetylsalicylic acid),[92]

naproxen[93] and sulindac are administered.[94]

The subjects most sensitive to acute renal impair-ment after NSAID use were those with ascitesand significant sodium retention. An importantreduction in natriuresis after the administrationof diuretics was observed in patients with asciteswho received two doses (with a 6-hour interval)of indomethacin 50 mg, naproxen 250 mg or as-pirin 900 mg. Urine sodium levels were 78% lower

in cirrhotic patients receiving furosemide (fruse-mide) and indomethacin than those in patientsreceiving only furosemide. The reduction in urinesodium levels was 49% for those receiving na-proxen and 17% for patients who took aspirin.This reduction did not occur or occurred to a verysmall extent in healthy subjects, confirming theincreased vulnerability of cirrhotic patients to theadverse effects of NSAIDs.[95] In the previousstudies, the decrease in natriuresis was reversed bydrug cessation. However, in these studies, NSAIDs

were usually administered for a very short time.Whether renal impairment is also reversible in cir-rhotic patients treated with NSAIDs for longerperiods is unknown.

Limited information is available on the useand effect of COX-2 selective inhibitors on renalfunction in cirrhotic patients. A double-blind,randomized, controlled study showed no appar-ent impairment in the renal function of patientswith cirrhosis and ascites when celecoxib was ad-ministered for 2.5 days. Neither the mean values

of GFR, renal plasma flow, and prostaglandin E2excretion nor the response to furosemide werereduced.[93] In a pilot study in nine cirrhotic

patients who received celecoxib for 4 days, nosignificant changes were observed in the meanvalues for serum creatinine, GFR, prostaglandinE2, urinary volume or sodium excretion before orafter drug administration. However, four patientsdisplayed a decrease in GFR greater than 20%.[96]

Experimental evidence of the expression of COX-2 in the kidney and their importance in renalhomeostasis were clearly established.[97] There-fore, it is difficult to imagine that COX-2 inhib-itors would present fewer renal problems thannon-selective NSAIDs in cirrhotic patients. These

findings, together with the decrease of GFR inseveral patients and the lack of studies of long-termuse, lead to the conclusion that the prescription of COX-2 inhibitors should be particularly restrictivein patients with hepatic disease.

The haemostatic abnormalities and coagula-tion disorders present in liver disease increasethe risk of bleeding in these patients.[98] One of the mechanisms leading to this coagulopathy isthe reduced platelet synthesis of proaggregatorythromboxane A2. NSAIDs inhibit the platelet

production of thromboxane A2, thus increasingthe risk of bleeding.[99]

Acute bleeding from oesophageal varices is amajor complication of hepatic cirrhosis. A case-control study found significant association be-tween the use of anti-inflammatory drugs and thefirst bleeding episodes associated with oesophagealor cardiac varices in cirrhotic patients. The studysuggested that cirrhotic patients using NSAIDsare approximately three times more likely topresent with this complication than are cirrhotic

patients who do not use these drugs.[100]

Owing toselective COX inhibition, the risk of acute bleed-ing from oesophageal varices might be lower if COX-2 inhibitors are used. However, the effect of these drugs on the first bleeding episodes asso-ciated with oesophageal or cardiac varices in cir-rhotic patients has not been investigated yet.

3.2.2 NSAID-Induced Liver Injury

Hepatotoxicity is considered a class character-istic of NSAIDs. Approximately 10% of all drug-

induced liver injuries are NSAID related.[101] Withmost NSAIDs, the mechanism of hepatic injury isconsidered idiosyncratic, dose independent and




10/25

dependent on individual susceptibility. An excep-tion is aspirin, which has intrinsic dose-dependenthepatotoxicity.[9] Although hepatotoxicity is listedas a class warning for NSAIDs, the risk of liverinjury differs among substances. NSAIDs such asbromfenac, ibufenac and benoxaprofen have beenwithdrawn from the market due to their hepato-toxicity. This serious adverse effect was also thereason for the withdrawal or lack of approvalof nimesulide and lumiracoxib in several coun-tries.[101,102] Higher drug-related hepatotoxicityhas also been observed with aspirin, diclofenac

and sulindac.

[103]

Although the risk of hepato-toxicity has not been evaluated in patients withunderlying liver disease, the use of these NSAIDsshould be avoided in this population.

3.2.3 Pharmacokinetics of Specific NSAIDs

in Hepatic Impairment

Most NSAIDs are eliminated via hepatic metab-olism involving oxidative (predominantly CYP2C9-catalysed) and conjugation reactions. The decreasedenzymatic activity in liver disease might result inmodification of the disposition of these drugs.Pharmacokinetic studies for several NSAIDs in he-patic impairment have been conducted in the pastdecades.

Aspirin

The pharmacokinetic properties of aspirin areunaffected in alcoholic patients with liver disease.However, the unbound fraction of its hydrolysedmetabolite, salicylic acid, is increased owing todecreased plasma protein binding. This decreaseresults in doubled AUC values of the free salicy-

late, indicating a higher risk for salicylate toxicityin these patients.[35]

Naproxen

Similarly, no differences in the total plasmaclearance of naproxen have been observed betweenindividuals with alcoholic cirrhosis and healthycontrols after single or multiple dose adminis-tration. The plasma protein binding of the drughas been significantly reduced in cirrhotic sub-

jects, resulting in a 2- to 4-fold increase of plasma

free drug concentration. A reduction of approxi-mately 60% has been observed for the unbounddrug clearance. Assuming that unbound drug

concentration determines pharmacological effect,naproxen doses in alcoholic cirrhotic patients shouldbe reduced by at least 50%.[36]

Ibuprofen

Pharmacokinetic studies with ibuprofen havesuggested that hepatic impairment has only aminimal effect on the disposition of the drug.Alcoholic liver disease had a small but not sta-tistically significant influence on the t½ and theAUC of ibuprofen.[37] Another study has dem-onstrated that the t½ is nearly doubled after theadministration of a single oral dose of ibuprofen

racemate.[104]

Etodolac

Despite the high protein binding and extensivehepatic metabolism of etodolac, no significantdifferences in the pharmacokinetics of this drughave been found in patients with stable cirrhosisand healthy volunteers after administration of asingle oral dose.[38]

Sulindac

Sulindac is a pro-drug, the bioactivation of

which leads to the active metabolite sulindacsulfide. One study showed that absorption wasdelayed in patients with poor hepatic function.The patients in the study displayed 3- and 4-foldincreases in the AUC for sulindac and sulindacsulfide, respectively, indicating the necessity fordose reduction of this drug in patients with he-patic impairment.[37]

Diclofenac

Diclofenac undergoes significant hepatic me-

tabolism and is highly protein bound. Thus, amodification in its pharmacokinetics might beexpected in the context of hepatic impairment.However, the pharmacokinetics of diclofenacwere unaffected after a single oral dose of diclo-fenac 100 mg in ten patients with chronic hepati-tis or compensated hepatic cirrhosis.[105] A morerecent study has demonstrated a 3-fold increasein the AUC in alcoholic cirrhotic patients but nochange in patients with chronic hepatitis com-pared with healthy subjects.[106] Because phar-

macodynamic measurements were not made andno increase in side effects was observed in thestudy, the authors suggested that doses should be




11/25

titrated to patient response instead of accordingto the severity of hepatic impairment.

Celecoxib

The pharmacokinetic properties of celecoxibare highly influenced by hepatic disease. A 22%increase in Cmax and a 40% increase in the AUChave been observed in patients with mild hepaticdisease. In patients with moderate hepatic dis-ease, the increases were 63% and 140% for Cmaxand AUC, respectively.[39] The metabolism rateis correlated with serum albumin levels. Half theusual dose is recommended in patients with

moderate hepatic disease (serum albumin levelsbetween 25 and 35 g/L). Studies in patients withsevere liver disease have not been conductedbecause celecoxib is contraindicated in thispopulation.[107]

Although the pharmacokinetic properties of certain NSAIDs appear to be unaltered in thepresence of mild to moderate liver disease, thesesubstances should be avoided in patients withadvanced liver disease owing to the increased riskof adverse effects. If used in patients with mild to

moderate liver disease, ibuprofen, etodolac anddiclofenac can be administered at normal doses,whereas dose reduction is necessary for napro-xen, sulindac and celecoxib.

3.3 Opioids

Opioids are largely used in the treatment of moderate to severe pain in a diverse patientpopulation. If used in patients with severe liverdisease or history of hepatic encephalopathy,

opioids may precipitate or aggravate encephalo-pathy.[108] This common and serious complicationin patients with severe liver disease is character-ized by abnormal mental status, ranging fromslight cognitive alterations to coma.[109] An increasein GABAergic inhibitory neurotransmission occursin hepatic encephalopathy. This increase has beenfound to decrease brain expression of proenke-phalin messenger RNA and thus decrease MET-enkephalin release. The decrease in endogenousopioid levels leads to compensatory up-regulation

of m-opioid receptors in the brain and increasedsensitivity to exogenous opioid analgesics.[110] Inaddition to these changes, alterations in the blood-

brain barrier in patients with severe liver diseasecan lead to increased drug concentrations withinthe central nervous system.[111]

Although the risk of precipitating encephalo-

pathy cannot be neglected, suitable pain manage-ment is important in patients with liver disease.When alternative analgesia is unavailable or in-sufficient, cautious use of opioids should beconsidered in these patients.[112,113] The pharm-acokinetics of these drugs in patients with hepaticimpairment are presented below to guide thechoice of suitable opioids. The major pathwaysand enzymes involved in the metabolism of eachopioid are shown in figure 2.

3.3.1 Codeine Codeine is a weak opioid analgesic chemically

related to morphine. It is metabolized by the liver

Opioid C Y P 2 B 6

C Y P 2 C 1 9

C Y P 2 D 6

C Y P 3 A 4

U G T

S u l f o t r a n s f e r a s e

E s t e r a s e s

Codeine

Tramadol

Tapentadol

Oxycodone

Morphine

Hydromorphone

Pethidine(meperidine)

Methadone

Buprenorphine

Fentanyl

Sufentanil

Alfentanil

Remifentanil

!

!

!

!

Main metabolic pathway

Minor metabolic pathway

Metabolic pathway leadingto active metabolite formation

!

Fig. 2. Major enzymes involved in opioid drug metabolism. CYP =cytochrome P450; UGT=uridine 50-diphosphate glucuronosyl-transferase.




12/25

mainly to codeine-6-glucuronide and norcode-ine, and a small fraction (approximately 10%) isO-demethylated to morphine.[114] Codeine itself hasa very weak affinity for m-opioid receptors.[115] Itsanalgesic activity is mainly due to the conversionto morphine, as several studies have demonstra-ted.[116-118] CYP2D6 is the enzyme implicated inthe biotransformation of codeine to morphine.As previously described, oxidative enzyme capa-city is reduced in patients with hepatic impair-ment. In this case, the result will be a reducedproduction of morphine and, in turn, a decrease

or lack of analgesia after codeine administration.Although a relative preservation of CYP2D6 ac-tivity may occur in mild liver disease, this preser-vation diminishes as impairment progresses.[23]

For instance, diminution of approximately 80%of the CYP2D6 metabolic activity is observed inchronic hepatitis C patients with liver kidneymicrosomal type 1 antibodies.[119]

Currently, no clinical studies demonstrate theanalgesic effect or the metabolism of codeine inpatients with hepatic impairment. Due to the lack

of studies and the possible lack of analgesic ef-fects, codeine appears to be a sub-optimal paintreatment choice in patients with liver disease.

3.3.2 Tramadol

More than 80% of tramadol is metabolized bythe liver.[120] The biotransformation of tramadolto its main metabolite, O-demethyl tramadol, iscatalysed by CYP2D6. Tramadol is characterizedby a bimodal mechanism of action: modulationof the central monoaminergic pathways and ac-

tivation of m-opioid receptors. Tramadol itself has higher monoaminergic activity, whereas itsmetabolite O-demethyl tramadol has higher affi-nity and activates m-opioid receptors more po-tently.[121-123] In patients with liver disease and,hence, lowered CYP2D6 activity, tramadol is ex-pected to act more as a monoaminergic mod-ulator than as an opioid agonist. One prospectiverandomized controlled study has shown that,compared with extensive CYP2D6 metabolisers,poor metabolisers consume more tramadol and

experience less postoperative pain relief.[124] Be-cause metabolizing capacity is reduced in patientswith liver disease, the analgesic effects of tram-

adol might be lower than expected in this groupof patients. However, this theory has not beendemonstrated in liver disease so far. Moreover,the monoaminergic effect of tramadol itself seemsto have analgesic effects because the pain toler-ance thresholds to sural nerve stimulation in poormetabolizers were significantly increased aftertramadol injection.[125] Significant differences wereobserved between healthy subjects and patientswith hepatic impairment in a study comparingthe pharmacokinetics of tramadol.[40] In patientswith liver cirrhosis, the AUC was increased by a

factor of 3.2 and the t½ by a factor of 2.6, onaverage. These changes are principally due to re-duced hepatic clearance. Similar changes in thepharmacokinetics of tramadol were observed inpatients with primary liver carcinoma on top of chronic hepatitis C. The bioavailability and AUCof tramadol were also increased but to a lesserextent in patients with secondary metastatic livermalignancy.[126] Owing to these metabolic changesand in order to prevent potential drug accumula-tion, prolonging the dosing intervals in patients

with hepatic impairment is recommended.

3.3.3 Tapentadol

Tapentadol is a new centrally acting analgesic,the mechanism of action of which is a combina-tion of m-opioid receptor agonism and inhibitionof noradrenaline (norepinephrine) reuptake.[127]

It undergoes important first-pass metabolism,explaining the bioavailability of only 32%. Thisanalgesic is extensively metabolized, mainly byconjugation to tapentadol-O-glucuronide (55%)

and tapentadol sulphate (15%).[41]

Phase I enzymesCYP2C9, CYP2C19 and CYP2D6 are responsiblefor 15% of the metabolism of tapentadol.

In a study conducted by the manufacturer,higher serum concentrations of tapentadol wereobserved in subjects with mild and moderate liverdisease than in subjects with normal hepaticfunction. AUC was increased by a factor of 1.7and 4.2 and t½ was increased by a factor of 1.2and 1.4 in subjects with mild and moderate liverdisease, respectively. These changes are probably

due to an increase in the bioavailability of thedrug. Although glucuronidation is somewhat pre-served in liver disease, the rate of formation of




13/25

tapentadol-O-glucuronide was lower in subjectswith increased liver impairment. It was suggestedthat no dose adjustment was necessary in patientswith mild liver disease. For patients with moder-ate liver disease, it was recommended that thetreatment should be initiated with the lowestavailable dose (50 mg) and increased dosing in-tervals (maximum of three doses in 24 hours). Nostudies were conducted in subjects with severe he-patic impairment and therefore the use of tapenta-dol is not recommended in this population.[41,128]

Currently, there are no recommendations on

the use of this substance in patients with hepa-torenal syndrome. More investigation and ex-perience in the use of this drug is needed in orderto confirm its safety in this population.

3.3.4 Oxycodone

Oxycodone is a semi-synthetic m-opioid ago-nist that has pharmacodynamic potency similarto that of morphine. Compared with morphine,oxycodone displays similar protein binding ca-pacity but a higher oral bioavailability (60 – 87%).

The metabolism of oxycodone depends on oxi-dative enzymes – notably CYP3A4 and CYP2D6 – which transform oxycodone to noroxycodoneand the active metabolite oxymorphone, respec-tively.[129] An impairment in oxycodone metabo-lism in liver disease might occur as a result of decreased hepatic blood flow and/or decreasedintrinsic clearance, since the metabolising activityof oxidative enzymes is reduced in chronic liverdisease. In this case, the formation of the activemetabolite oxymorphone would be reduced, lead-

ing to potentially lower analgesic effects as ob-served in poor CYP2D6 metabolizers.[130,131]

In patients with hepatic impairment, the Cmaxof oxycodone was increased by 40% and the AUCby 90% after administration of a 20 mg control-led-release oxycodone tablet. Reductions of 15%of the Cmax and 50% of the AUC of the activemetabolite oxymorphone also occurred. The t½of oxycodone was prolonged by 2 hours.[46] Thesedata suggest that oral oxycodone should be in-itiated at lower doses in patients with hepatic

impairment.Important differences in pharmacokinetic pa-

rameters have been observed before and after liver

transplantation in end-stage cirrhotic patientswhen oxycodone was administered intravenously.The median t½ was 13.9 hours (range 4.6 – 24.4hours) before transplantation and 3.4 hours (range2.6 – 5.1 hours) after transplantation, and the clear-ance increased from 0.26 L/min before to 1.13L/minafter transplantation. In these patients, higherventilatory depression was observed before trans-plantation, which was believed to be the result of the increased sensitivity to the adverse effects of opioids in cirrhotic patients.[47] Because of theimportant increase in median t½ and AUC, the

dosage interval of oxycodone should be increasedand/or doses should be reduced in patients withsevere liver cirrhosis.

3.3.5 Morphine

Morphine undergoes significant first-pass me-tabolism after oral administration, and its aver-age bioavailability is 30 – 40%. The drug is weaklybound to plasma proteins (20 – 40%).[132] Metabolismof morphine to the active morphine-6-glucuronideand the inactive but neurotoxic morphine-3-glu-

curonide occurs mainly in the liver.[132]

Morphineis an intermediately to highly extracted drug witha hepatic extraction ratio of approximately 0.7.[133]

Hence, the possible decreased clearance of mor-phine in cirrhotic patients should be mostly dueto a decrease in hepatic blood flow and, to asmaller extent, a decrease in intrinsic metaboliz-ing capacity. Although the plasma protein bind-ing of morphine is decreased in hepatic disease,[134]

the higher amount of the free fraction is expectedto have no significant impact on the Vd because

morphine is only weakly protein bound.Several studies have investigated the disposi-

tion of morphine in patients with hepatic impair-ment. Patwardhan et al.[135] found no significantalteration in morphine elimination and plasmaclearance in cirrhotic patients (Child-Pugh A or B)after intravenous administration. In contrast, fewother studies have shown impairment in the me-tabolism of intravenous morphine in patientswith liver disease.[42-44] In a study by Mazoitet al.,[44] the terminal t½ in cirrhotic patients was

2-fold greater and the clearance decreased by 37%compared with that in normal subjects. The au-thors suggest that the dosing interval of morphine




14/25

should be increased 1.5- to 2-fold in cirrhoticpatients in order to avoid accumulation of thedrug. The change in these pharmacokinetic pa-rameters was even more pronounced in a studyby Hasselstro ¨ m et al.[43] that included patientswith Child-Pugh B or C hepatic impairment. Crottyat al.[42] have reported a reduction of 25% in theextraction ratio of morphine in cirrhotic patients.They conclude that this reduction is due to di-minished intrinsic hepatic clearance (reduction inthe enzyme activity or intrahepatic shunting) be-cause no differences in hepatic blood flow were

observed. Their study also demonstrated thatsystemic clearance was significantly higher thanhepatic clearance, furnishing some indirect sup-port for the possible extra-hepatic metabolism of morphine. The extrahepatic conjugases found inthe kidneys or intestines might assume a greaterrole in morphine elimination in liver failure.[19]

The differences in morphine elimination valuesin the study by Patwardhan et al.[135] comparedwith those in other studies[42-44] are principallydue to differences in the severity of the liver dis-

ease of the subjects studied. As mentioned insection 2.2, glucuronidation, which is relativelypreserved in mild to moderate liver disease, mightbe impaired in severe liver disease.[20]

As a result of decreased first-pass metabolism,the oral bioavailability of morphine in patients withhepatic impairment is likely to be increased. Thishas been demonstrated in a study by Hasselstro ¨ met al.[43] in which the oral bioavailability of mor-phine in cirrhotic patients was 100% comparedwith 47% in control subjects after a single oral

dose. In another study, the bioavailability afteradministration of controlled-release morphinetablets to cirrhotic patients was 27.7%, whereasthat in controls was 16%.[45] These studies alsoshowed prolongation of the t½ and a decrease inmorphine clearance in cirrhotic patients.

An important increase in the bioavailability of controlled-release morphine was also noted in astudy of patients with liver carcinoma. Bioavaila-bility was 64.8% in patients with primary liver car-cinoma, 62.1% in patients with secondary metastatic

carcinoma, and 16.8% in controls. Consequently,the AUC was increased 4-fold in primary carci-noma and 3-fold in metastatic carcinoma.[136]

The data presented above indicate that if morphine is given intravenously to patients withsevere liver disease, the dosage interval should beincreased. In the case of oral administration, notonly should the administration interval be pro-longed but the dose should also be reduced.Morphine should be avoided in patients withhepatorenal syndrome because of the increasedrisk of neurotoxicity resulting from morphine-3-glucuronide and morphine 6-glucuronide accu-mulation in severe renal impairment.

3.3.6 Hydromorphone

Hydromorphone is a semi-synthetic opioidthat undergoes important first-pass metabolism,resulting in low oral bioavailability.[137] It is pre-dominantly metabolized by glucuroconjugationto hydromorphone-3-glucuronide. Several othermetabolites are formed in smaller amounts: hydro-morphone-3-glucoside, dihydromorphine, and un-conjugated and conjugated dihydroisomorphine.

In patients with moderate hepatic impair-ment, Cmax and AUC were increased 4-fold after

single-dose administration of oral immediate-releasehydromorphone. This increase was probably aconsequence of reduced first-pass metabolism.The t½ of the drug in patients with hepatic im-pairment was the same as that in controls.[48]

According to the results, a reduction of hydro-morphone dose with maintenance of the standarddosing interval is necessary in patients with mod-erate liver disease. Possible decreases in the meta-bolizing capacity of conjugating enzymes withthe advancement of liver disease may lead to an

increase in the t½ in patients with severe liverdisease. However, no studies investigating thepharmacokinetics of hydromorphone in patientswith severe liver disease are currently beingundertaken. In the presence of renal impairment,an accumulation of the neuroexcitatory metabo-lite hydromorphone-3-glucuronide has been ob-served.[138,139] Therefore, hydromorphone shouldbe avoided in patients with hepatorenal syndrome.

3.3.7 Pethidine (Meperidine)

Pethidine (meperidine) was the first syntheticopioid analgesic.[140] It is predominantly meta-bolized by hydrolysis to meperidinic acid, which




15/25

is conjugated and excreted, but it is alsoN-demethylated by CYP3A4 to normeperidine(norpethidine). This metabolite has neurotoxiceffects and has been implicated in the develop-ment of neuromuscular irritability and sei-zures.[141,142] The oral bioavailability of pethidineis approximately 50%.[49,50] Thus, to obtainequianalgesia, oral doses should be twice as highas intravenous doses, generating plasma con-centrations of normeperidine that are higher afteroral than those after intravenous administration.

In cirrhotic patients, a 60 – 80% increase in bio-

availability was observed after oral administra-tion.[49,50] Significant impairment in pethidinedisposition also occurred after intravenous ad-ministration, with a decrease of approximately50% in the plasma clearance and a 2-fold increasein the t½.

[50,51] The decreased formation of nor-pethidine in cirrhotic patients might lead to theconclusion that these patients are relatively pro-tected from its toxicity. However, the slowerelimination of the metabolite might lead to anincreased risk of cumulative toxicity if repeated

doses are administered.[50,52]

In conclusion, if administered to patients with hepatic impair-ment, oral doses of pethidine should be reduced.Repeated doses of pethidine should be avoidedbecause of the risk of norpethidine accumulationand neurotoxicity. Further accumulation of nor-pethidine occurs in patients with renal impair-ment; thus, this analgesic should be avoided inpatients with hepatorenal syndrome.[142]

3.3.8 Methadone

Methadone is a synthetic opioid predominantlyused as maintenance treatment in individuals withopioid dependence. It has high average bioavaila-bility of approximately 70 – 80%, but large variabil-ity has been reported (36 – 100%).[143] The proteinbinding of methadone is high (60 – 90%). It is mainlybound toa1-acid glycoprotein, and its distribution isnot significantly altered by hypoalbuminaemia.[143]

Methadone is metabolized by oxidation, withprincipal involvement of CYP3A4 and CYP2B6.The elimination of methadone and its metabolites

is urinary and faecal.[144]As previously mentioned, methadone is lar-

gely used as maintenance treatment in patients

with chronic opioid addiction. A significant pro-portion of these patients have chronic hepatitis Cthat can progress to cirrhosis. The prevalence of hepatitis C virus antibodies in patients enrolledin methadone maintenance programmes is veryhigh, ranging from 67% to 96%.[145,146] In a studyof 14 methadone-maintenance patients, the dis-position of methadone was unaltered in subjectswith mild to moderate chronic liver disease. Inpatients with more severe liver disease, the t½ wasprolonged from 19 to 35 hours, but drug clear-ance and AUC were not significantly altered.[53]

Similar results have been observed in patientswith severe alcoholic cirrhosis,[54] leading to thesuggestion that no dose adjustment is necessaryin these patients. Moreover, a CYP3A4 inductionhas been suggested as an explanation for the re-quirement of higher doses of methadone in pa-tients with hepatitis C.[147] The use of methadonefor analgesia and not as maintenance treatmentin patients with hepatic impairment has not beeninvestigated. Methadone disposition seems to berelatively unaffected in renal impairment;[148]

thus, its clearance should not be decreased fur-ther in the presence of hepatorenal syndrome.However, due to the important interindividualvariability in the pharmacokinetics of methadoneas well as its long t½, this drug should not be usedas a first-line analgesic treatment in patients withliver disease.

3.3.9 Buprenorphine

Buprenorphine is a partial agonist at the m-opioidreceptors. It has very high first-pass clearance

and is therefore not administered orally but onlyusing sublingual, parenteral or transdermal routes.The bioavailability of sublingually administeredbuprenorphine is 50 – 55%, with important inter-individual variability.[149,150] This drug is highlyprotein bound (96%), primarily to a- and b-glo-bulin.[151] Buprenorphine is partially metabolizedby the liver, with the main metabolic pathway beingoxidation to norbuprenorphine by CYP3A4.[152]

Both buprenorphine and norbuprenorphine arefurther glucuronidated.[153] The elimination is

mainly through the faeces (80 –

90%), mostly asfree buprenorphine and norbuprenorphine. Theremaining 10 – 20% is eliminated in the urine as




16/25

metabolites.[154,155] Enterohepatic recirculationprobably occurs, resulting in an apparent pro-longation of the t½.

[151]

Buprenorphine pharmacokinetics were notstudied in patients with hepatic impairment.Sublingually administered buprenorphine theo-retically bypasses the liver; however, the drugmight be partially swallowed and thus subjectedto hepatic first-pass metabolism, which mightexplain the average bioavailability (50 – 55%) andlarge variability. Decreased CYP3A4 enzymaticactivity in liver disease might result in an increase

of the bioavailability and decrease of the clear-ance of buprenorphine. However, owing to thepartial buprenorphine metabolism and the partialbypass of the liver with the sublingual adminis-tration, these changes might be of low clinicalrelevance.

Several cases of buprenorphine hepatic toxi-city have been described, most frequently afterintravenous use of the drug.[156-158] Contra-dictory results exist regarding the hepatotoxicityof buprenorphine in patients already presenting

liver disease, particularly hepatitis C. One studydemonstrated elevated transaminases in patientswith a history of hepatitis who were treated withtherapeutic doses of sublingual buprenorphine.The increase in aspartate aminotransferase (AST)was dependent on buprenorphine dose.[159] Noevidence of buprenorphine hepatotoxicity wasfound in another study that included adolescentsand young adults, among whom 19% were he-patitis C positive.[160] Safe use of buprenorphinein patients with active hepatitis C has been sug-

gested in a case series study in which no increasewas observed in the transaminases of four patientstreated with buprenorphine.[161] The authors of this study have suggested that patients with he-patitis C should not be excluded from treatmentwith buprenorphine. A more prudent course,however, would be to monitor liver enzyme levelscarefully if buprenorphine is administered to thisgroup of patients. Buprenorphine appears to be asafe option for pain treatment in patients withrenal disease.[162,163] Although the pharmaco-

kinetics of the drug might be relatively unchangedin liver disease, no studies confirming this hypo-thesis are currently available.

3.3.10 Fentanyl

Fentanyl is a synthetic opioid from the phenyl-piperidine class. Similar to the other drugs of thisclass, it exhibits multiple-compartment pharmaco-kinetics. It is highly protein bound (85%), mainlyto albumin. Fentanyl is largely metabolized in theliver by CYP3A4. Its t½ is approximately 3.6 hours,with large interindividual variability. Importantprolongation of the t½ was observed in patientsreceiving continuous infusion of fentanyl.[164]

Since the hepatic extraction ratio of fentanyl ishigh (0.8), its clearance would mainly be affected

by changes in hepatic blood flow, not by a re-duction in intrinsic enzyme activity or proteinbinding.[165]

The pharmacokinetics of fentanyl were unalteredin patients with biopsy-confirmed cirrhosis after asingle intravenous dose of fentanyl.[55] However,none of these patients had profound hepatic in-sufficiency, and their hepatic blood flow was notmarkedly diminished compared with that in healthysubjects. These results should be interpreted withcaution if fentanyl is administered to patients

with hepatic shunting or reduced hepatic bloodflow.

The pharmacokinetics of transdermal fentanylmatrix patches in cirrhotic patients have beenstudied by their manufacturer. The Cmax andAUC were increased by 35% and 73%, respectively,and the t½ remained unchanged after application of a fentanyl matrix patch (50mg/hour).[166]

The pharmacokinetics of continuously infusedfentanyl in cirrhotic patients have not been est-ablished, and whether the accumulation of fen-

tanyl is more pronounced in these patients than inpatients with normal liver function is unknown.Fentanyl has often been reported as a first-choiceopioid in renal impairment,[167-169] although itsclearance might be reduced in the presence of high blood urea nitrogen levels.[170] This opioidappears to be a good choice in patients with he-patorenal syndrome, but dose reduction might benecessary to avoid accumulation, especially withcontinuous administration.

3.3.11 Sufentanil Sufentanil is another drug in the piperidine

opioid class. Compared with fentanyl, it is more




17/25

lipophilic, but has a slightly smaller Vd and shortert½

. It is highly protein bound (92%), mainly toa1-acid glycoprotein. Sufentanil is extensivelymetabolized by CYP3A4 in the liver and has ahepatic extraction ratio approaching 1.[164]

Similar to that of fentanyl, the pharmaco-kinetics of sufentanil are not influenced by liverdisease after a single intravenous dose.[56] The pro-posed explanations for the unaffected pharm-acokinetics in patients with mild liver disease area possible sparing of liver blood flow or the in-capacity to detect the differences in elimination

kinetics owing to the large Vd. A 30%

prolonga-tion of the t½, slight increase in the Vd, and in-crease in the clearance in cirrhotic patients havebeen reported by the manufacturer.[171]

Like that of fentanyl, the t½ of continuouslyinfused sufentanil is increased in patients withnormal liver function.[164] No studies have beenperformed to determine the degree of possibleaccumulation of continuously infused sufentanilin patients with hepatic disease. Sufentanil pharm-acokinetics are not significantly altered in renal

impairment,[164]

and this opioid, like fentanyl,may be used in patients with hepatorenal syndrome.

3.3.12 Alfentanil

Alfentanil is a short-acting opioid that has arapid onset but an analgesic effect that lasts nolonger than 5 – 10 minutes. It has a significantlysmaller Vd and shorter t½ than fentanyl and su-fentanil. The a1-acid glycoprotein binding of al-fentanil is approximately 92%.[164] It is extensivelyand almost exclusively metabolized by the CYP3A

enzymes.[172]

Owing to the intermediate hepaticextraction ratio of alfentanil (0.3 – 0.6),[57,164] itstotal hepatic clearance could be influenced by all of the following: hepatic blood flow, intrinsic hepaticenzyme activity and protein binding.

A substantial increase in the t½ (219 vs 90 min-utes), and a 50% decrease in total clearance havebeen reported in patients with moderate liverdisease. Moreover, protein binding decreasedfrom 88.5% to 81.4%. When corrected to proteinbinding, a decrease of 70% in the plasma clear-

ance of the unbound fraction has been observedin hepatically impaired patients.[57] Another studyhas shown similar results for alfentanil disposi-

tion in anaesthetised patients with hepatic pa-thology.[58] The disposition of alfentanil in childrenwith cholestatic hepatic disease was found to beunaltered.[173] The discrepancy between the resultsof this study in children and those of previousstudies might be due to differences in underlyingpathology or patient age. Moreover, the length of plasma sampling in this study was only 2 hours,which might explain the lack of detection of thepotential pharmacokinetic alterations.

A more recent study, in which a 3-fold increasein AUC in patients with mild liver cirrhosis was

observed, confirmed the important alterations of alfentanil disposition even in patients with minorhepatic impairment.[59] Thus, alfentanil seems tobe a poor analgesic choice in patients with liverdisease because its effects may be both prolongedand enhanced.

3.3.13 Remifentanil

Remifentanil is a phenylpiperidine opioid,which differs considerably from other opioids inits class because of its ester linkages that lead to a

specific metabolic pathway. As an ester, remifentanilis predominantly and rapidly hydrolysed by bloodand tissue esterases to a carboxylic acid metabolite,which has been found to have only 1/4600 of theparent compound potency in animal models.[174,175]

This particular metabolic pathway explains its veryshort duration of action and rapid elimination.

A study of the pharmacokinetic parametersof remifentanil demonstrated no change after a4-hour infusion in subjects with severe hepaticimpairment. Patients with liver disease seemed to

be more sensitive to the ventilatory depressanteffects of remifentanil. Owing to the short dura-tion of action of this drug, the increased sensi-tivity in this population is unlikely to have clinicalsignificance.[60]

The clearance of remifentanil in the anhepaticphase of liver transplantation is similar to that of healthy volunteers, confirming the extrahepaticmetabolism of the drug and its independencefrom hepatic function.[61] The pharmacokineticsof remifentanil in patients with renal failure are

also unaltered.[176] These results suggest that doseadjustment is unnecessary in patients with liverdisease or hepatorenal syndrome.




18/25

3.4 Neuropathic Pain Treatment

Neuropathic pain is a medical challenge as it ispoorly responsive to classical anti-inflammatoryor powerful centrally acting analgesics, such asopioids.[177] Evidence-based guidelines suggestthe use of antidepressants and anticonvulsants asfirst-line neuropathic pain treatment.[178,179] Stud-ies evaluating the disposition, safety and efficacy of neuropathic pain drugs in patients with hepaticimpairment are often lacking. In this population,alternative, non-pharmacological interventionsshould be encouraged whenever possible. How-

ever, in some cases when neuropathic pain is notsufficiently relieved by non-pharmacological inter-ventions, drug administration could be considered.

3.4.1 Antidepressants

Tricyclic Antidepressants

Several randomized controlled clinical studieshave demonstrated the efficacy of tricyclic anti-depressants (TCAs) as neuropathic pain treat-ment.[180] These drugs are largely metabolized

by liver cytochromes, CYP2D6 in particular. Inpatients with liver disease where a decrease of cytochrome activity is expected, an accumulationof these drugs is possible. In this population,treatment with TCAs should be started at lowdoses with slow titration. Nortriptyline and de-sipramine could offer the same efficacy andshould be preferred over amitriptyline and imip-ramine when available since they appear to be lesssedating and better tolerated.

Serotonin-Norepinephrine Reuptake InhibitorsThe use of serotonin-norepinephrine reuptake

inhibitors (SNRIs) such as venlafaxine and du-loxetine for the treatment of neuropathic pain isincreasing. Venlafaxine undergoes significant hepa-tic biotransformation to several inactive and oneactive metabolite mediated primarily by CYP2D6and to a lesser extent by CYP3A4. In patients withmoderate hepatic impairment, significant altera-tions were observed in the t½ (30% and 60%prolongation) and clearance (50% and 30% de-

crease) of venlafaxine and its active metabolite,respectively. In patients with severe hepatic im-pairment, a decrease of up to 90% was observed

in venlafaxine clearance. An important inter-individual variability exists, making dosage adjust-ments difficult in this population.[181] Significantalterations in the disposition of duloxetine (85%clearance decrease) were observed in patientswith moderate liver disease.[182] Moreover, du-loxetine hepatotoxicity has been evidenced. Patientswith pre-existing liver disease appear to be at higherrisk of duloxetine-induced liver injury. These find-ings prompted the manufacturer to include awarning in the product label stating that dulox-etine ‘‘should ordinarily not be prescribed to a

patient with substantial alcohol use or evidence of chronic liver disease.’’[183]

Selective Serotonin Reuptake Inhibitors

Selective serotonin reuptake inhibitors (SSRIs)show lower efficacy than TCAs in the treatmentof neuropathic pain.[177] Moreover, these drugsmight increase the risk of gastrointestinal bleed-ing from varices in patients with hepatic impair-ment and thus are not the first-choice treatmentfor neuropathic pain in this population.[184]

3.4.2 Anticonvulsants

Calcium Channel a2d Ligands

Anticonvulsants are the second drug classlargely used in the treatment of neuropathic pain.Among them, calcium channel a2d ligands suchas gabapentin and pregabalin are currently oftenused as first-line medications. The disposition of gabapentin and pregabalin is probably unalteredin patients with hepatic impairment since bothdrugs are excreted renally without previous metab-

olism and are not bound to plasma proteins.[185]

Moreover, gabapentin was not found to be clearlyassociated with hepatic injury, and thus probablyrepresents the safest choice for the treatment of neuropathic pain in patients with liver disease.Several cases of pregabalin hepatotoxicity havebeen reported. In one of the reported cases,pregabalin hepatotoxicity occurred in a patientwith underlying liver disease.[186] Physicians mustbe aware of the possible, although rare, occur-rence of pregabalin-induced or -aggravated liver

injury. As for other patients, in the case of hepa-tic impairment, these drugs should be startedat low doses and titrated cautiously in order




19/25

to minimize the dose-dependent dizziness andsedation.[187]

Other Anticonvulsants

Other anticonvulsants used in the treatment of neuropathic pain such as carbamazepine are con-traindicated in patients with hepatic impairmentdue to the important risk of hepatotoxicity.[176]

3.4.3 Opioids

Neuropathic pain states were, for a long time,considered resistant to opioid analgesia. How-ever, some randomized controlled trails have

shown a decrease in neuropathic pain after opioidtreatment.[188] The use of opioids in patients withhepatic impairment is discussed in section 3.3.Due to the potential risk of development of tol-erance or addiction with the long-term use of these drugs and the risk of aggravating hepaticencephalopathy, opioids should be used cau-tiously and only as second- or third-line neuro-pathic pain treatment in this population.

4. Conclusion and Clinical

Recommendations

The fear of aggravating pre-existing liver diseaseoften leads to undertreatment of pain in patientswith hepatic impairment. Ideally, analgesics as wellas other hepatically cleared or hepatotoxic drugs,should be avoided in this population and non-pharmacological interventions should be preferredwhenever possible. However, in some situations,such as postoperative pain, avoiding the use of analgesics would be unethical. Analgesic drugs

can be used in patients with hepatic impairment,but the choice of drug and its dose must be madecarefully.

In the limited number of studies existing onthis subject, the very young and very elderly po-pulations have often been left out. The selectionof suitable drug or dose is even more difficult forthis extreme age population or for patients withother co-morbidities.

Drug-drug interactions are another concern inpatients with hepatic impairment. For example,

the co-administration of NSAIDs with otherdrugs that could provoke gastrointestinal bleed-ing, such as low-dose aspirin or SSRIs, or with

drugs that could impair glomerular filtration, suchas angiotensin-converting enzyme inhibitors, shouldbe avoided in this particularly vulnerable popula-tion. Furthermore, opioids should not be com-bined with any other sedative or anxiolytic drugsto reduce the risk of precipitating hepatic en-cephalopathy. From a pharmacokinetic point of view, in this population, physicians should avoidthe prescription of drugs altering CYP activitywhich can further modify the metabolism andelimination of other hepatically cleared drugs,analgesics included.

When choosing an analgesic, physicians shouldfollow the guidelines for the type of pain and thenselect an analgesic within these guidelines thatwould be suitable and safe in patients with he-patic impairment.

Paracetamol at low doses (maximum 2 g/day)and for a short duration can be used in patientswith hepatic impairment for the treatment of weaknociceptive pain. When paracetamol is prescribed,informing the patient about the maximal dailydose and the presence of this drug in many over-

the-counter medications is important. NSAIDsshould be avoided in patients with liver diseasebecause of their antiplatelet activity, gastro-intestinal irritation, the increased risk of acuterenal failure, and the potential and unpredictablerisk of drug-induced liver injury (e.g. with diclo-fenac, lumiracoxib and nimesulide).

The disposition of most opioids is affected insevere liver disease. The efficacy of some of them,such as codeine and possibly tramadol and oxy-codone, might be compromised because their

biotransformation to active opioids is decreased.Other opioids, such as pethidine, should be avoi-ded because of possible accumulation of toxicmetabolites.

When using opioids in patients with hepaticimpairment, the dosing regimen should be care-fully established. For highly extracted drugs, suchas morphine or hydromorphone, the initial oraldose must be reduced owing to increased bio-availability. For drugs with decreased clearance,repeated doses should be decreased, or dosing in-

tervals increased in order to avoid drug accumu-lation. The best approach in hepatically impairedpatients is individual titration with short-acting




20/25

opioids to achieve optimal doses for pain relief without significant adverse effects.

Glucuroconjugated morphine or hydromor-phone at reduced doses, and intravenous fenta-nyl, sufentanil and remifentanil appear to be thebest opioid choices in patients with liver disease.

Opioids such as morphine, pethidine or hydro-morphone, which have renally cleared active ortoxic metabolites, should be avoided in the pre-sence of hepatorenal syndrome. The dispositionsof phenylpiperidine opioids – fentanyl, sufentaniland remifentanil – appear to be unaffected by

hepatorenal syndrome. From a theoretical pointof view, buprenorphine might be a potential opioidto use in patients with liver disease. However, ad-ditional clinical studies are needed to provide evi-dence of its disposition and safety in this group of patients. Further research is also necessary to de-termine the disposition of continuously adminis-tered fentanyl and sufentanil and the analgesic ef-fects of codeine and tramadol in patients with liverdisease.

All patients with hepatic impairment receiving

opioids must be carefully monitored for any signsof hepatic encephalopathy, regardless of themedication prescribed.

Gabapentin or low-dose TCAs appear to bethe safest options for the management of neuro-pathic pain in patients with hepatic impairment.

Acknowledgements

The preparation of this review was supported by a grantfrom the Swiss National Science Foundation (NK-23K1-122264). The authors have no conflicts of interest that aredirectly relevant to the content of this review.

References1. Morgan DJ, McLean AJ. Clinical pharmacokinetic and

pharmacodynamic considerations in patients with liverdisease: an update. Clin Pharmacokinet 1995 Nov; 29 (5):370-91

2. Larrey D, Pageaux G. Prescribing drugs in liver disease. In:Rodes J, Benhamou J, Blei A, editors. Textbook of he-patology: from basic science to clinical practice. Malden(MA): Blackwell Publishing, 2007: 1912-22

3. Verbeeck RK. Pharmacokinetics and dosage adjustment in

patients with hepatic dysfunction. Eur J Clin Pharmacol2008 Dec; 64 (12): 1147-61

4. US Department of Health and Human Services, Food andDrug Administration, Center for Drug Evaluation and

Research, Center for Biologics Evaluation and Research.Guidance for industry: pharmacokinetics in patients with

impaired hepatic function. Study design, data analysis,and impact on dosing and labeling. 2003 May [online].Available from URL: http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM072123.pdf [Accessed 2011 Dec 19]

5. European Medicines Agency, Committee for MedicinalProducts for Human Use. Guideline on the evaluation of the pharmacokinetics of medicinal products in patientswith impaired hepatic function. London: 2005 Feb 17[online]. Available from URL: http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/09/WC500003122.pdf [Accessed 2011 Dec 19]

6. Hebert MF. Guide to drug dosage in hepatic disease:Avery’s drug treatment. A guide to the properties, choice,

therapeutic use and economic value of drugs in diseasemanagement. 4th ed. Auckland: Adis International, 1997:1761-89

7. Lee WM. Drug-induced hepatotoxicity. N Engl J Med 1995Oct 26; 333 (17): 1118-27

8. Larson AM, Polson J, Fontana RJ, et al. Acetaminophen-induced acute liver failure: results of a United Statesmulticenter, prospective study. Hepatology 2005 Dec; 42(6): 1364-72

9. Tolman KG. Hepatotoxicity of non-narcotic analgesics.Am J Med 1998 Jul 27; 105 (1B): 13S-9S

10. Lee CH, Wang JD, Chen PC. Increased risk of hospitali-zation for acute hepatitis in patients with previous ex-

posure to NSAIDs. Pharmacoepidemiol Drug Saf 2010Jul; 19 (7): 708-14

11. Traversa G, Bianchi C, Da Cas R, et al. Cohort study of hepatotoxicity associated with nimesulide and other non-steroidal anti-inflammatory drugs. BMJ 2003 Jul 5; 327(7405): 18-22

12. Laine L, White WB, Rostom A, et al. COX-2 selective in-hibitors in the treatment of osteoarthritis. Semin ArthritisRheum 2008 Dec; 38 (3): 165-87

13. Radner H, Ramiro S, Buchbinder R, et al. Pain managementfor inflammatory arthritis (rheumatoid arthritis, psoriaticarthritis, ankylosing spondylitis and other spondylarthritis)and gastrointestinal or liver comorbidity. Cochrane Data-

base Syst Rev 2012; 1: CD008951

14. Rossi S, Assis DN, Awsare M, et al. Useof over-the-counteranalgesics in patients with

Documents

Analgesia in Liver Impair Pts