Comparison of potentially hepatotoxic drugs among major US drug compendia

Research in Social and Administrative Pharmacy 1 (2005) 460–479

Comparison of potentially hepatotoxic drugs among major US drug compendiaqqq Jeff J. Guo, Ph.D.a,b,*, Patricia R. Wigle, Pharm.D.a, Kangsan Lammers, Pharm.D.c, Oceana Vu, Pharm.D.d a

University of Cincinnati College of Pharmacy, University of Cincinnati Medical Center, 3225 Eden Avenue, Cincinnati, OH 45267-0004, USA b Institute for Health Policy and Health Services Research, University of Cincinnati, Cincinnati, OH 45267 USA c Children Hospital, University of Cincinnati Medical Center, Cincinnati, OH 45237 USA d Merck-Medco Pharmacy Services, Cincinnati, OH 45246 USA

Abstract Background: Although a large number of drugs include warnings or listed adverse reactions that describe reports of associated hepatotoxicity, the hepatotoxic risk is documented with different definitions in major drug compendia. Objectives: The purposes of this study were to compare inclusion of potentially hepatotoxic drugs, and analyze the ratings of hepatotoxic risk among major drug compendia. Methods: To assess the risk of drug-associated hepatotoxicity, we used current literature of epidemiological studies and developed a 4-level rating scale of hepatotoxic drugs: 3, clear literature evidence of life-threatening hepatotoxicity; 2, multiple case reports or qqq The authors have no conflict of interest concerning this research. We did not receive any consulting fees, grants, honoraria, patents, royalties, stocks, or other financial or material gain that may involve the subject matter of the research. This study was presented at the International Conference of Pharmacoepidemiology, Bordeaux, France, August 22–25, 2004. * Corresponding author. University of Cincinnati College of Pharmacy, University of Cincinnati Medical Center, 3225 Eden Avenue, Cincinnati, OH 45267-0004, USA. Tel.: C1 513 558 8613; fax: C1 513 558 4372. E-mail address: jeff[email protected] (J.J. Guo).

1551-7411/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.sapharm.2005.06.005

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

461

significant liver injuries; 1, no significant liver damage has been reported; and 0, no information. All drugs were evaluated using the 5 major US drug compendia: American Hospital Formulary Service (AHFS ), United States Pharmacopeia Drug Information (USPDI ), Facts and Comparisons (F&C ), Physicians’ Desk Reference (PDR), and Clinical Pharmacology (CP). Average rating scores were calculated as the sum of each drug rating score divided by the total number of drugs. One-way analysis of variance and independent t tests were conducted to compare the difference among the rating scores. Results: In total, 175 different drugs and 3 therapeutic classes with hepatotoxic effects were identified in the compendia, including 59 antineoplastics, 28 antiinfectives, 17 nonsteroidal anti-inflammatory drugs, 17 antipsychotics or phenothiazine derivatives, 9 angiotensin-converting enzyme inhibitors, 6 anticonvulsants, 4 histamine-2 receptor antagonists, and other drugs. Average rating scores were 1.65 for AHFS, 1.10 for USPDI, 1.27 for F&C, 1.34 for PDR, and 1.61 for CP (F Z 7.93, P ! .0001). The risk categories were significantly different among compendia in 4 therapeutic classes of antipsychotics and/or phenothiazines (F Z 3.471, P Z .011), nonsteroidal anti-inflammatory drugs (F Z 7.866, P ! .0001), antineoplastics (F Z 2.476, P Z .044), anti-infectives (F Z 2.003, P Z .098), and angiotensinconverting enzyme inhibitors (F Z 38.125, P ! .0001). Conclusions: Rating scores of hepatotoxicity were significantly different among drug compendia. The different compendium put different emphasis on hepatotoxicity severity. Comprehensive evaluations of hepatotoxic-related drugs provide critical information for health practitioners. Ó 2005 Elsevier Inc. All rights reserved. Keywords: Hepatotoxicity; Risk Category; Rating Score; Drug Compendium

1. Introduction Drug-induced liver disease is a relatively common, but often unrecognized cause of liver injury. Liver injury can be associated with several therapeutic classes of medications; most commonly antituberculosis agents, antibiotics, antiretroviral, and nonsteroid anti-inflammatory drugs (NSAIDs).1-7 Severe hepatotoxicity is one of the most common causes for pharmaceutical product recalls, labeling changes, and ‘‘Dear Doctor Letter’’ interventions required by the US Food and Drug Administration. Adverse hepatic reactions accounted for 18% of postmarketing withdrawals in France, Germany, the United Kingdom, and the United States between 1961 and 1992 and have continued to be the leading cause of such withdrawals in the last few years.8 Hepatic injury accounts for 3.5%-9.5% of all adverse drug reaction reports and up to 14.7% of fatal adverse reactions.9 Using data from the computerized Danish National Hospital Registry, the incidence of drug-induced liver disease was calculated as 19 and 22 per million person-years for men and women, respectively.10 In the United States, Galan et al11 reported that 1% of all hepatology patients presented with drug-induced hepatitis. Based on the UK General Practice Research Database, the relative risks of drug-induced idiopathic or acute liver

462

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

injury for 40 drugs/therapeutic classes were calculated.12 Those drugs included chlorpromazine, amoxicillin/clavulanic acid, flucloxacillin, macrolides, tetracyclines, metoclopramide, and chlorpheniramine, to name but some. Drug-induced hepatic injury is sometimes predictable but also may be idiosyncratic.13 Serious liver disorder caused by drugs usually can be characterized as hepatocellular necrosis, cholestatis injury, or a combination of both. Drug-associated acute hepatocellular necrosis is characterized by elevated serum transaminase levels and is often clinically indistinguishable from hepatocellular necrosis of viral, ischemic, or other etiologies. Hepatocellular necrosis has been associated with acetaminophen, isoniazid, and other drugs.14,15 Although a very large number of drugs include warnings or listed adverse reactions that describe reports of associated hepatotoxicity, the actual incidence of drug-induced liver injury is not well understood for overall drugs and/or specific drug groups. Drug-induced hepatic injury accounts for more than 50% of the cases of acute liver failure.1 Based on the US Acute Liver Failure Group Study,16 acetaminophen overdose was the most common apparent cause of acute liver failure accounting for 39% of cases; idiosyncratic drug reactions were the presumptive cause in 13% of cases; viral hepatitis A and B combined were implicated in 12% of cases, and 17% of cases were of indeterminate cause.8 Drugs are responsible for 2%-6% of jaundice and about 10% of cases of ‘‘acute hepatitis.’’17,18 Drug-induced chronic hepatitis has been considered rare, even though it accounts for up to 6% of all chronic hepatitis.19 Despite considerable progress in toxicological studies, the correlation in liver toxicity between animals and man remains poor.20 Following is a typical scenario, along with subsequent questions/ considerations, that might be encountered in clinical practice settings: a patient presents to a busy clinician’s office to receive the results of his lab work. The liver function tests are noted to be elevated.  Was this adverse event caused by a medication the patient is taking? Has the medication been associated with more harmful hepatic adverse events? To find the answers to these questions, the clinician can select from a number of electronic, text, and personal data assistant references. B If the suspect medication is not listed in a table of potentially hepatotoxic medications in one reference, is it cleared of having caused potential injury? B If the suspect medication is listed as causing clinically significant liver dysfunction in one reference, is that a true assessment of the medication’s hepatotoxic nature?  What are the criteria that made this event clinically significant?  Was it an elevation in liver function tests greater than 3 times the upper limit of normal?  Was it the presence of jaundice? Was it a fatality?  Does the absence of a black box warning mean the drug does not have clinically significant hepatotoxicity in its past? Searching the primary

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

463

literature may or may not provide any clear cut information than those previously mentioned. Whether the resulting wealth of information, ranging from case reports to review articles, about various drugs provides more questions than answers is debatable. Two of the biggest problems facing clinicians and pharmacoepidemiologists about drug-induced hepatotoxicity are lack of agreement on the most appropriate sources for accurate information and standardized criteria to evaluate the hepatotoxic risk. In clinical practice, awareness of drug-induced hepatotoxicity is very important for clinical practitioners because early recognition is critical. When assessing the risk of liver injury associated with a particular drug (eg, troglitazone, bromfenac, trovafloxacine), pharmacoepidemiologists have to control the concomitant medications that are definitely and/or probably associated with hepatotoxicity. It is difficult for a researcher to choose a ‘‘gold-standard’’ drug compendium because each drug compendium is unique in its comprehensives and presentation of dose, adverse effects, kinetics, and pharmacology information. Indeed, considerable amount of data are available about how drugs can cause hepatotoxicity (mechanism), when these effects tend to occur (onset, recurrence with rechallenge), and the type of injury caused by the drug (acute versus chronic, reversible versus irreversible, location). However, there is still a lack of standardized criteria with which to evaluate the hepatotoxic nature of a drug. There is no grading scale like that for hepatic encephalopathy or the Child Pugh scores that determines how hepatotoxic a particular medicine is or could be. Because a large number of drugs are documented as hepatotoxic with different definitions of hepatotoxicity in major drug compendia, we proposed to compare and analyze major drug compendia with respect to inclusion of potentially hepatotoxic drugs and to compare the rating of emphasis attached to the hepatotoxicity among these drug compendia.

2. Methods 2.1. Risk category of hepatotoxic drugs Frequency of hepatic adverse reactions for some drugs is well documented in the medical literature.21–26 For the purposes of this study, a list of potentially hepatotoxic drugs was generated from the US Food and Drug Administration adverse drug spontaneous reports from 1990 to 1997 and published literature. These adverse drug reports were largely collected from manufacturers, clinical practitioners, and patients via the MedWatch program. The initial list contained 121 drugs that caused hepatotoxicity. Primary literature was also used to develop a general risk category of hepatotoxicity. The approximate estimates of hepatotoxic risk category have been based upon the following 4 rating levels:

464

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

3, black box warning drugs or clear evidence from the literature of recognized life-threatening hepatotoxicity (eg, isoniazid, troglitazone); 2, multiple cases of hepatic failure or significant liver injuries in the literature; 1, no significant liver damage has been reported in literature; and 0, no information about hepatotoxicity was mentioned (see Figure 1 and Table 1). Figure 1 shows that the risk of hepatotoxicity increased with higher rating scores from 0 to 3. Obviously, this classification is not mutually exclusive and has overlapping risk categories. The overlapping circles may be due to a paucity of published data used for development of drug compendia. With more population-based pharmacoepidemiologic research, the overlapping circles of risk categories will be less and the difference will be more obvious. The average rating scores were calculated as the sum of each drug rating score divided by the total number of drugs. We calculated the average rating scores for overall hepatotoxic drugs and 5 major therapeutic classes: antipsychotics/phenothiazines, NSAID, antineoplastics, anti-infectives, and angiotensin-converting enzyme (ACE) inhibitors. 2.2. Data sources and evaluation process Sources of drug compendia included American Hospital Formulary Service Drug Information (AHFS ) 2003, United States Pharmacopeia Drug Information (USPDI ) 2004, Facts and Comparisons (F&C ) 2003, Physicians’ Desk Reference (PDR) 2003, and Clinical Pharmacology (CP) 2003. These 5 drug compendia are commonly used by physicians, pharmacists, nurses, and other health care practitioners in the United States. All drug monographs were reviewed based on their electronic versions. We focused on some specific sections, such as precautions to consider, cautions, adverse effects, adverse reactions, administration, other effects, gastrointestinal effects, side

Group 0

Group 1

Group 2

Increasing Risk of Hepatotoxicity

Figure 1. Risk categories of hepatotoxic drugs.

Group 3

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

465

Table 1 Risk of categories for hepatotoxic drugs Category levela Category description 0

No information about hepatotoxicity.

1

No significant liver damage has been reported.

2

Multiple case reports or significant liver injuries.

3

A clear literature evidence of life-threatening hepatotoxicity or death. Drug may also have a black box warning due to medicationinduced hepatotoxicity.

Evaluation criteria and standardsb  Hepatotoxicity, hepatic injury, or hepatic damage was not mentioned in any section of monograph.  No relevant data were reported.  Data from animal sources; no human case reports  Or, drug was not found in the compendium.  Extremely rare case reports of significant liver damage were mentioned.  Asymptomatic liver injury.  Liver function or liver enzyme elevation (eg, ALP, AST, LDH, ALT tests) was mentioned, and was reversible.  Abnormal liver function or hepatic dysfunction was mentioned, including fever, biliary stasis, jaundice, yellowing of the eyes or skin, abdominal pain, but was reversible.  No incidence or prevalence data reported.  Several case reports of significant liver damage were mentioned.  Liver function or liver enzyme elevation (eg, ALP, AST, LDH, ALT tests) was mentioned, and may not be reversible.  Human cases of hepatitis or hepatic necrosis were mentioned.  Abnormal liver function or hepatic dysfunction was mentioned, including fever, biliary stasis, jaundice, yellow eye, skin, abdominal pain, but may not be reversible.  Incidence or prevalence data were mentioned.  Life-threatening hepatotoxicity.

 Fatalities reported.  Black box-warnings noted.  Documented cases of liver failure or liver transplantation.  Warnings of severe hepatic necrosis or drug-induced hepatotoxicity were mentioned. a Risk category of hepatotoxicity increases from level 0 to level 3. Each level may not be mutually exclusive. b Abbreviations: ALP, alkaline phosphatase; AST, aspartate aminotransferase; LDH, lactate dehydrogenase; ALT, alanine aminotransferase.

466

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

effects, hepatic effects and contraindications. Key words used in the search primarily included hepatotoxicity, hepatotoxic, hepatic, hepatic necrosis, hepatocellular, steatosis, liver, liver toxicity, liver enzyme elevation, liver function, jaundice, liver damage, liver injury, liver failure, liver transplantation, and drug-induced hepatotoxicity. For each related drug, we documented compendium name, drug brand name, generic name, therapeutic class, black box warning (yes or no), rating score, original monograph statements related to drug hepatotoxicity, and comments related to coding that the evaluator might have deemed necessary in a predesigned Microsoft Access database. Two junior pharmacists (newly trained and licensed) reviewed each drug monograph and its hepatotoxic risk and copied/pasted relevant information into the Access database. Then, they evaluated and provided the draft ratings and comments. These 2 pharmacists (one of whom work in a hospital, the other, in a mail-order pharmacy) were identified because of previous interest in hepatotoxicity research during a clinical clerkship experience. After the data were collected, 2 different pharmacists (a Pharm.D. and Ph.D.) with several years of clinical practice and research experience re-evaluated and confirmed the risk categories of hepatotoxic drugs. The latter 2 pharmacists were clinical/ research faculty members who were advisors for both junior pharmacists during their clinical training. Of 178 risk rating scores in 5 compendia, 85.5% ratings were consistent between 2 junior pharmacists and the other 2 pharmacists (Pearson correlation r Z 0.867). The other 14.5% disagreement of drug rating scores was double-checked based on the original drug monographs, and the latter 2 pharmacists arrived at a consensus. Here are examples of how the developed rating scale was applied for 2 specific hepatotoxic drugs. Ranitidine (ZantacÒ)  AHFS: category 3. ‘‘.Hepatitis, which may be hepatocellular, hepatocanalicular, or both and may or may not be accompanied by jaundice, has occurred occasionally . However, death has been reported rarely . Hepatic failure has been reported rarely.’’  USPDI: category 3. ‘‘.Rarely deaths have occurred; Rare cases of hepatic failure have been reported, as have rare cases of cholestatic or mixed hepatocellular and cholestatic liver toxicity.’’  F&C: category 1. ‘‘.Occasionally, reversible hepatitis, hepatocellular or hepatocanalicular or mixed, with or without jaundice has occurred with oral ranitidine alanine aminotransferase values have increased to at least twice pretreatment levels with IV ranitidine administered for 5 days.’’  PDR: category 3. ‘‘There have been occasional reports of hepatocellular, cholestatic, or mixed hepatitis, with or without jaundice . Rare circumstances death has occurred . Rare cases of hepatic failure have also been reported.’’  CP: category 1. ‘‘.Rare hepatitis, jaundice, and elevated hepatic enzymes have been reported with ranitidine.’’

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

467

Acetaminophen (TylenolÒ)  AHFS: category 3. ‘‘Hepatic necrosis may occur . Hepatic coma is likely to develop if the plasma half-life of acetaminophen exceeds 12 h. Fulminant fatal hepatic failure may occur in chronic alcoholics following over dosage of acetaminophen . Fatalities are rare with less than 15 g .’’  USPDI: category 1. ‘‘Risk of hepatotoxicity with single toxic doses or prolonged use of high doses of acetaminophen may be increased in alcoholics or in patients regularly taking other hepatotoxic medications or hepatic enzyme inducers . Hepatic function determinations may be required at periodic intervals during high-dose or long-term therapy, especially in patients with pre-existing hepatic disease.’’  F&C: category 2. ‘‘Hepatotoxicity and severe hepatic failure occurred in chronic alcoholics following therapeutic doses. The hepatotoxicity is believed to be caused by induction of hepatic microsomal enzymes resulting in an increase in toxic metabolites or by the reduced amount of glutathione responsible for conjugating toxic metabolites . The minimal toxic dose is 10 g (140 mg/kg), but liver damage has occurred with a single 5.85 g dose .’’  PDR: category 3. ‘‘Combined with heavy drinking, the acetaminophen in these products could conceivably cause liver damage . In adults and adolescents ( R 12 years), hepatic toxicity may occur following ingestion of greater than 7.5 to 10 g over a period of 8 h or less. Fatalities are infrequent (less than 3%-4% of untreated cases) and have rarely been reported with overdoses of less than 15 g.’’  CP: category 2. ‘‘In most cases, hepatotoxicity occurs as a result of an acute overdose, however, moderately excessive doses, if taken chronically, can also produce hepatotoxicity. Acetaminophen-induced hepatotoxicity is manifested as hepatic necrosis, jaundice, bleeding, and encephalopathy.’’ 2.3. Statistical analysis Mean rating scores for hepatotoxic drugs were calculated for all study drugs and major subgroups of hepatotoxic drugs. One-way analysis of variance was conducted to compare the difference of rating scores among the 5 drug compendia. An independent-sample t test was used to compare the difference of rating scores between each pair of drug compendia. Alpha level of statistical significance was predetermined as .05. All statistical analyses were conducted using Statistical Package for Social Sciences (SPSS) for Windows, version 10. 3. Results Table 2 summarizes ratings and inclusion of 175 drugs and 3 therapeutic classes with hepatotoxic effects among 5 major compendia, including 59

468

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

Table 2 Ratings and inclusion of hepatotoxic drugs in 5 drug compendia (n Z 175) Category generic name

Rating and inclusionb Brand namea

AHFS

USP

F&C

PDR

CP

Antipsychotics or phenothiazine derivativesc Olanzapine Zyprexa Thioridazine Mellaril Triflupromazine Vesprin Risperidone Risperdal Perphenazine Trilafon Fluphenazine Prolixin Chlorpromazine Thorazine Acetophenazined Tindal Ethopropazine Pardidol Methotrimeprazine Levoprome Prochlorperazine Compazine Promazine Sparine Promethazine Phenergan Propiomazine Largon Thiethylperazine Torecan Trimeprazine Temaril Clozapine Clozaril

2 3 0 2 3 3 3 0 0 0 3 0 3 0 0 0 2

1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 1

1 1 1 1 0 1 1 0 0 0 0 0 1 0 0 1 1

2 1 0 2 0 0 0 0 0 0 1 0 1 0 0 0 2

2 1 0 1 1 1 1 0 0 0 1 0 1 0 1 0 0

NSAIDs Bromfenacd Diclofenac Diflunisal Etodolac Fenoprofen Ibuprofen Indomethacin Ketoprofen Ketorolac Meclofenamate Nabumetone Naproxen

0 3 2 0 3 3 3 2 2 2 0 3

0 1 0 0 1 1 1 1 1 1 1 1

0 0 0 1 1 1 1 1 1 1 1 2

0 3 2 0 0 0 3 0 2 0 3 3

3 3 2 2 3 2 3 2 2 2 2 3

0 0 2 3 3

1 1 0 1 1

1 0 0 2 1

0 0 0 2 0

2 0 2 2 3

Antineoplastics and biological response modulators Azathioprine Imuran 3 Estramustine Emcyt 2 Thioguanine Lanvis, Tabloid 1 Trimetrexate Neutrexin 0 Aldesleukin Proleukin 3 Anastrozole Arimidex 0 Asparaginase Elspar 3 Bicalutamide Casodex 0

1 1 1 1 1 0 1 1

3 1 1 1 3 1 3 3

2 0 1 0 3 2 1 3

3 2 2 1 2 0 1 2

Oxaprozin Phenylbutazone Piroxicam Sulindac Tolmetin

Duract Cataflam, Voltaren Dolobid Lodine Nalfon Advil, Motrin Indocin Orudis KT, Oruvail Toradol Meclomen Relafen Aleve, Anaprox, Naprosyn, Naprelan Daypro Butazolidine Feldene Clinoril Tolectin

469

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479 Table 2 (continued) Category generic name Bleomycin sulfate Busulfan Capecitabine Carboplatin Carmustine Chlorambucil Cisplatin Cladribine Cyclophosphamide Cytarabine Dacarbazine Dactinomycin Daunorubicin Docetaxel Doxorubicin Epirubicin Etoposide Exemestane Floxuridine Fludarabine Fluorouracil Flutamide Gemcitabine Hydroxyurea Ifosfamide Interferon alfa Irinotecan Letrozole Leuprolide acetate Lomustine Mechlorethamine Megestrol acetate Melphalan Mercaptopurine Methotrexate Plicamycin Mitomycin Mitotane Mitoxantrone Nilutamide Paclitaxel Pegaspargase Pentostatin Procarbazine Rituximab Streptozocin Tamoxifen citrate

Rating and inclusionb Brand namea

AHFS USP F&C PDR CP

Blenoxane 0 Busulfex, Myleran 3 Xeloda 2 Paraplatin 2 BiCNU, Gliadel 3 Leukeran 1 Platinol, Platinol-AQ 1 Leustatin 1 Cytoxan, Neosar 1 DepoCyt 1 DTIC-Dome 3 Cosmegen 2 DaunoXome 1 Taxotere 3 Adriamycin 1 Ellence 0 Toposar, VePesid 1 Aromasin 0 FUDR 1 Fludara 2 Adrucil, 5-FU 0 Eulexin 0 Gemzar 0 Droxia, Hydrea 0 Ifex 2 Alferon N, Intron A, 2 Roferon-A, Wellferon Camptosar 0 Femara 0 Lupron, Eligard, Viadur 1 CeeNU 1 Mustargen 0 Megace 0 Alkeran 1 Purinethol, 6-MP 3 Rheumatrex 3 Mithracin, Mithramycin 0 Mutamycin 0 Lysodren 1 Novantrone 0 Nilandron 3 Taxol, Onxol 3 Oncaspar, PEG2 L-asparaginase Nipent, 2#-deoxycoformycin, DCF 2 Matulane 1 Rituxan 0 Zanosar 3 NOLVADEX, Tamofen 3

1 1 2 0 1 1 0 0 0 0 3 3 1 0 1 0 1 0 1 2 0 1 0 0 1 1

1 1 2 1 1 0 1 1 1 1 3 1 1 2 2 2 2 1 0 0 0 3 2 1 1 3

0 3 2 2 0 0 0 1 0 0 3 1 0 2 0 0 0 0 0 2 0 0 3 0 0 3

0 2 2 2 3 1 1 0 2 1 3 2 0 3 0 0 0 0 1 2 2 3 3 1 2 3

1 1 0 1 1 0 0 1 2 0 0 0 1 3 0 2

1 1 1 1 0 0 1 3 0 0 0 1 0 3 1 1

2 0 0 0 1 0 2 3 3 1 0 0 2 3 0 2

2 2 0 1 0 1 2 3 3 2 0 0 1 2 2 1

1 2 2 2 1 1 1 1 0 0 3 0 1 2 0 3 3 3 2 3 (continued on next page)

470

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

Table 2 (continued) Category generic name

Rating and inclusionb Brand namea

AHFS USP F&C PDR CP

Temodar Vumon, VM-26 Fareston Herceptin

0 0 0 2

0 1 1 0

1 1 1 0

0 0 2 2

0 0 1 0

Antimigraine agents Eletriptan Sumatriptan

Relpax Imitrex

0 1

0 0

1 1

0 1

0 0

Antihypertensive Methyldopa Spironolactone

Aldomet Aldactone

3 0

3 0

3 0

3 0

2 0

Anticonvulsants Primidone Carbamazepine Felbamate Phenytoin Succinimidee Valproic acid

Mysoline Tegretol Felbatol Dilantin Celontin, Zarontin Depakene, Depakote

0 2 3 0 0 3

1 1 2 2 0 3

0 0 3 1 1 3

0 2 2 2 0 3

2 2 3 2 1 1

Antituberculosis/anti-infectivef Isoniazid Nydrazid Pyrazinamide Tebrazid Rifampin Rifadin

3 3 3

3 1 1

3 1 3

3 3 3

2 3 3

Antifungal/anti-infective Fluconazole Itraconazole Ketoconazole Terbinafine Griseofulvin

Diflucan Sporanox Nizoral Lamisil Fulvicin, Grifulvin, Gris-PEG

3 3 3 0 1

3 1 3 2 1

3 3 3 3 1

3 3 3 3 0

3 3 3 3 2

Antimalarial/anti-infective Chloroquine Hydroxychloroquine

Aralen Plaquenil

0 0

1 0

1 1

0 1

0 1

Antiviral/anti-infective Efavirenz Ritonavir Saquinavir

Sustiva, EFV Norvir Fortovase, Invirase

2 3 2

1

1

2 3 2

2 3 2

Antibiotics/anti-infective Amoxicillin and clavulanate Azithromycin Ceftriaxone Cilastatin Clarithromycin

Augmentin 3 Zithromax 3 Rocephin 2 Primaxin (with imipenem) 2 Biaxin 3

1 1 1 0 2

0 1 0 1 0

3 3 2 2 3

1 3 1 0 3

Temozolomide Teniposide Toremifene Trastuzumab

471

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479 Table 2 (continued) Category generic name

Rating and inclusionb Brand namea

AHFS USP F&C PDR CP

EryTab, E.E.S. Floxapen Primaxin (with cilastatin) Dynacin, Minocin Macrobid, Macrodantin Bicillin, Pfizerpen, Wycillin Sulfisoxazole diolamine Bactrim, Septra Trovan Sumycin

1 0 2 3 3 1 1 1 3 3

2 2 0 1 2 1 1 1 3 1

1 0 1 1 3 0 0 1 3 3

2 0 2 1 3 0 0 2 3 0

2 0 0 3 3 0 3 1 2 2

ACE inhibitors Benazepril Captopril Enalapril Fosinopril sodium Lisinopril Moexipril Quinapril Ramipril Trandolapril

Lotensin Capoten Vasotec Monopril Prinivil, Zestril Univasc Accupril Altace Mavik

3 3 3 3 3 3 3 3 3

2 2 2 2 2 2 2 2 2

2 0 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3

3 3 3 3 3 3 3 3 3

Histamine-2 receptor antagonists Cimetidine Ranitidine Famotidine Nizatidine

Tagamet Zantac Pepcid Axid

3 3 1 2

3 3 3 3

1 1 1 1

3 3 1 1

2 1 0 2

Muscle relaxant Chlorzoxazone Dantrolene

Parafon Forte DSC Dantrium

3 3

1 1

1 3

0 3

2 3

Protopic, Prograf Imuran Gengraf, Neoral, Sandimmune

2 3 2

1 0 1

0 0 2

2 3 2

2 3 2

Disease-modifying antirheumatic drugs Aurothioglucose Solganal Penicillamine Cuprimine, Depen Leflunomide Arava

3 2 3

1 1 1

1 1 2

0 2 3

0 1 3

Other categories Acetaminophen Etretinated Acitretin Anabolic steroidse

3 0 0 0

1 1 1 1

2 0 2 3

3 0 3 0

2 0 3 1

Erythromycin Flucloxacillin Imipenem Minocycline Nitrofurantoin Penicillin Sulfonamidese Trimethoprim-sulfamethoxazole Trovafloxacin Tetracycline

Immunosuppressant agent Tacrolimus Azathioprine Cyclosporine

Tylenol Tegison Soriatane Anadrol, Oxandrin, Dianobol, Winstrol

(continued on next page)

472

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

Table 2 (continued) Category generic name Allopurinol Estrogens Tiopronin Metoclopramide Pemoline Propylthiouracil Sulfasalazine Tolcapone Ondansetron Trimethobenzamide Charcoal Naltrexone Troglitazoned Nefazodone Bupropion Tocainide Isotretinoin Fluothane Fluoxymesterone Disulfiram Donepezil Tacrine

Rating and inclusionb Brand namea Zyloprim Premarin Thiola Reglan Cylert PTU Azulfidine Tasmar Zofran Tigan CharcoCaps ReVia Rezulin Serzone Wellbutrin, Wellbutrin SR, Zyban Tonocard Accutane Halothane Halotestin Antabuse Aricept Cognex

AHFS

USP

F&C

PDR

CP

2 3 0 1 3 3 1 3 2 1 0 2 0 3 2

1 1 1 0 3 2 1 3 1 1 0 1 2 3 0

1 1 1 0 3 2 1 3 0 1 0 2 3 3 1

2 1 0 0 3 0 2 3 2 1 0 0 0 2 2

2 1 0 1 3 3 3 3 2 1 0 2 3 3 0

2 2 0 1 3 0 2

0 1 1 1 3 0 2

1 2 2 3 3 0 2

0 2 0 0 3 0 0

0 2 3 1 3 0 2

Abbreviations: AHFS, American Hospital Formulary Society; USPDI, United States Pharmacopeia; F&C, Facts & Comparisons; PDR, Physicians’ Desk Reference; CP, Clinical Pharmacology Drug Information. a Not all brand names were listed. Trade Mark Ô or register Ò was omitted. b Rating scores (risk categories) of hepatotoxicity: 0, 1, 2, and 3 as defined in Table 2. c The AHFS provided the hepatotoxic information of phenothiazine derivatives as a group effect. d Discontinued, withdrawal, or not available in US market. e Related to one drug therapeutic class. f The PDR provided the hepatotoxic information of antituberculosis as a group effect.

(33%) antineoplastics, 28 (15%) anti-infectives (including 15 antibiotics, 5 antifungal drugs, 3 antiretroviral agents, and 3 antituberculosis drugs), 17 (10%) NSAID, 17 (10%) antipsychotics or phenothiazines, 9 (5%) ACE inhibitors, 6 (3%) anticonvulsants, and 4 (2%) histamine-2 receptor antagonists. Of these, 3 drug therapeutic classes were evaluated by group (1) succinimide (eg, ethosuximide, methsuximide), (2) sulfonamide (eg, sulfasalazine, sulfisoxazole, sulfanilamide, sulfamethoxazole), and (3) anabolic steroids (eg, oxymetholone, oxandrolone, methandrostenolone, stanozolol). These drugs were evaluated as a group because their hepatotoxic effects were reported as a class effect in the drug compendia. Four drugs (bromfenac, troglitazone, acetophenazine, etretinate), which were withdrawn or not available in US market were also included.

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

473

We found 148 potentially hepatotoxic drugs in AHFS, 174 drugs in USPDI, 168 drugs in F&C, 146 drugs in CP, and only 119 drugs in PDR based on our criteria of risk categories from 0 to 3 in Table 1. With relevant hepatotoxic information (risk categories 1, 2, and 3), we identified 127 (85.6%) drugs in AHFS, 136 (77.8%) drugs in USPDI, 136 (80.6%) drugs in F&C, 104 (87.2%) drugs in PDR, and 136 (93.1%) drugs in CP (see Table 2). Average rating scores were 1.65 for ASHP, 1.10 for USPDI, 1.27 for F&C, 1.34 for PDR, and 1.61 for CP (see Table 3). For overall hepatotoxic drugs, the mean scores of risk categories among 5 compendia were statistically and significantly different (F Z 7.928, P ! .0001). The risk categories of hepatotoxic drugs were also compared among 5 compendia for major therapeutic groups as follows:  The risk categories for antipsychotics or phenothiazines were significantly different among compendia (F Z 3.471, P Z .011). The average AHFS rating score (1.41) was higher than those of other compendia.  The risk categories for NSAIDs were significantly different among compendia (F Z 7.866, P ! .0001). The rating scores for CP and AHFS were higher than those of other compendia.  The risk categories for antineoplastics and biological response modulators were significantly different among compendia (F Z 2.476, P Z .044).  The risk categories for anti-infective drugs (including antibiotics, antifungal, antimalarial, antituberculosis, and antivirus) were not significantly different among compendia (F Z 2.003, P Z .098).  The risk categories for ACE inhibitors were significantly different among compendia (F Z 38.125, P ! .0001). Based on independent t tests of risk categories for hepatotoxic drugs between each pair of drug compendia (see Table 4), AHFS drug information had significantly higher ratings than USPDI (t Z 4.877, P ! .0001), F&C (t Z 3.177, P Z .002), and PDR (t Z 2.321, P Z .021); USPDI rating scores were significantly lower than PDR (t Z ÿ2.142, P Z .033) and CP (t Z ÿ4.837, P ! .0001); the rating scores from both F&C and PDR were significantly lower than those from CP (P Z .003 and P Z .034, respectively).

4. Discussion The present study examined and compared the ratings of risk category and inclusion of hepatotoxic drugs among 5 major US drug compendia. It is the first content analysis study to compare and rate potentially hepatotoxic drugs because no other studies were found in the literature. The US Pharmacopeia Drug Information (USPDI ), Facts and Comparisons, and Clinical Pharmacology documented more total numbers of hepatotoxic drugs than the other 2 compendia American Hospital Formulary Society Drug Information (AHFS ) and Physician Desk Reference (PDR). These

474

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

Table 3 Analysis of variance of risk categories for hepatotoxic drugs among 5 drug compendia (N Z 176) and their subcategories 95% Confidence interval Upper bound

N

Overall hepatotoxic drugs AHFS

178 1.65

1.25 1.47

1.84

178 1.10

0.87 0.97

1.23

1.27 1.34 1.61 1.39

1.01 1.27 1.16 1.13

1.12 1.16 1.45 1.32

1.42 1.53 1.78 1.47

1.41

1.42 0.68

2.14

USPDI

17 0.82

0.39 0.62

1.03

Facts and comparisons Physicians’ desk reference Clinical pharmacology Total

17 17 17 85

0.52 0.80 0.62 0.88

0.27 0.12 0.27 0.59

0.79 0.94 0.91 0.97

17 1.82

1.29 1.16

2.48

USPDI

17 0.76

0.44 0.54

0.99

Facts and comparisons Physicians’ desk reference Clinical pharmacology Total

17 17 17 85

0.64 1.34 0.75 1.11

0.50 0.37 1.85 1.10

1.15 1.75 2.62 1.58

Antineoplastics and biological response modulators AHFS 59 1.25 1.17 0.95

1.56

USPDI

Facts and comparisons 178 Physicians’ desk reference 178 Clinical pharmacology 178 Total 890 Antipsychotics or phenothiazines AHFS 17

NSAID AHFS

USPDI

Mean SD

Lower bound

Drug compendia

0.53 0.53 0.59 0.78

0.82 1.06 2.24 1.34

59 0.83

0.83 0.61

1.05

0.98 1.21 1.07 1.07

1.02 0.84 1.13 1.06

1.53 1.47 1.69 1.31

28 2.04

1.14 1.59

2.48

USPDI

28 1.36

0.87 1.02

1.69

Facts and comparisons

28 1.54

1.20 1.07

2.00

Facts and comparisons 59 Physicians’ desk reference 59 Clinical pharmacology 59 Total 295 Anti-infective drugs AHFS

1.27 1.15 1.41 1.18

Mean squares

F ratio (P value)

Between groups 7.928 9.847 Within groups (!.0001) 1.242

Between groups 3.471 2.394 Within groups (.011) 0.690

Between groups 7.866 7.276 Within groups (!.0001) 0.925

Between groups 2.476 2.775 Within groups (.044) 1.121

Between groups 2.003 2.511 Within groups (.098) 1.253

475

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479 Table 3 (continued ) 95% Confidence interval Drug compendia

N

Mean SD

Physicians’ desk reference 28 1.96 Clinical pharmacology 28 1.93 Total 140 1.76 ACE inhibitors AHFS USPDI Facts and comparisons Physicians’ desk reference Clinical pharmacology Total

Lower bound

Upper bound

1.20 1.50 1.15 1.48 1.14 1.57

2.43 2.38 1.95

9 3.00

0.00 3.00

3.00

9 2.00

0.00 2.00

2.00

0.67 0.00 0.00 0.62

2.29 3.00 3.00 2.74

9 9 9 45

1.78 3.00 3.00 2.56

1.27 3.00 3.00 2.37

Mean squares

F ratio (P value)

Between groups 38.125 3.389 Within groups (!.0001) 0.089

references also provided relatively more detailed hepatotoxic drug information from clinical trials and postmarketing epidemiologic evaluations. For health care practitioners, it is more likely to find hepatotoxic-related drugs using these 3 compendia. These drug compendia are different in type of texts, ie, PDR is based on manufacturer’s product label information; AHFS provides a very comprehensive review of adverse effects but no references; Facts and Comparisons provides a summary table of drug adverse effects; USPDI contains quite comprehensive review and original literature references, but it seems to provide very conservative hepatotoxic evaluation for most of drugs. As for the lower number of hepatotoxic drugs documented in PDR, we believe that the primary reason could be that the PDR provides primarily brand name products currently available in the market. For these products, the hepatotoxic data may be limited to premarketing clinical trial information provided by manufacturers. PDR excludes classes of medications, discontinued medications, and medications Table 4 Matrix of independent t tests of risk categories for hepatotoxic drugs between each pair of drug compendia Drug compendia USPDI Facts and comparisons Physicians’ desk reference Clinical pharmacology

AHFS

USPDI

Facts and comparisons

Physicians’ desk reference

4.877 (P ! .0001) 3.177 (P Z .002) ÿ1.738 (P Z .083) 2.321 (P Z .021)

ÿ2.142 (P Z .033)

ÿ0.601 (P Z .548)

0.313 (P Z .755)

ÿ4.838 (P ! .0001) ÿ3.021 (P Z .003) ÿ2.123 (P Z .034)

476

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

not available in the United States. Unfortunately, US doctors may be more likely to consult the PDR than other compendia, given free access to online resources and/or a lack of availability or familiarity with references more commonly used by pharmacists. This means that doctors may miss how high the hepatotoxic risk is for a particular drug. Our findings indicate that the different compendia place different emphasis on hepatotoxicity severity. For example, Clinical Pharmacology and AHFS provided higher risk categories for NSAIDs than other compendia. The AHFS rating scores were higher than those of other compendia for antipsychotics and/or phenothiazines. AHFS, F&C, and CP provided higher risk categories for antineoplastics and biological response modulators than other compendia. AHFS, CP, and PDR provided higher risk categories for ACE inhibitors than other compendia. USPDI provided relatively lower rating scores in almost all categories. Hepatic injury is one of the most common forms of preclinical and clinical toxicity encountered with drugs. Because the liver is central to the biotransformation of virtually all drugs and foreign substances, drug-induced liver injury is a potential complication of nearly every medication that is prescribed. Because preapproval studies are neither designed nor are they able to detect rare, serious events such as severe hepatotoxicity, it is critically important that new hepatotoxic information appearing during postmarketing surveillance be collected and analyzed in an expeditious manner. In the recent experiences of severe hepatotoxicity observed during postmarketing surveillance with troglitazone and bromfenac, there have been suggestions of less severe liver effects (eg, elevated transaminases) in the premarketing clinical trials. Hence, major drug compendia, like AHFS, USPDI, and F&C, provided with comprehensive premarketing and postmarketing clinical evaluation of hepatotoxic information are critical and useful for clinical practitioners. On the other hand, the monitoring for drug-induced hepatotoxicity is an important responsibility that should be shared by the regulatory authorities, drug manufacturers, physicians, pharmacists, and other clinical practitioners. Their timely reporting of adverse drug events contributes to everyone’s understanding about rare adverse events. The risk category of hepatotoxic drugs in this study is limited by many factors. Three drug therapeutic classes (succinimide, sulfonamide, and anabolic steroids) were evaluated as the class effect in these compendia. It is often clinically challenging to determine whether an adverse effect is a class effect or caused by an individual drug. One of the more significant is the absence of consideration of dose. This is particularly true regarding overthe-counter medicines and dietary supplements. It accounts for the largest number of drug-associated liver transplants in the United States, but most of them can be used safely and effectively when used within recommended dosing guidelines. The present study did not review herbals and over-thecounter medications because the study is limited to comparing inclusion and ratings among 5 major US compendia. The original references that

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

477

construct the drug monographs were not compared because it would be difficult and is beyond the scope of this article to attempt to establish standardized criteria for drug-induced hepatotoxicity. With information technology development, many clinicians use innovative drug information systems, eg, Micormedex & Epocrates, for their data sources because traditional drug compendia become dated so rapidly. Because of limited resources, the electronic systems and foreign drug compendia, such as Martindale were not reviewed. In the present study, the content analysis of hepatotoxicity risk categories was primarily conducted by 4 different pharmacists based on drug monographs in 5 compendia. The evaluation did not rate or take into consideration patient factors, disease states, or case reports. The criteria used were determined based on different levels of risk to the patient. The criteria did not differentiate based on mechanism of injury because drug-induced hepatotoxicity can be multifactorial and separating out the mechanism(s) and location(s) of injury may prove difficult, if not impossible while evaluating the electronic resources. The onset and recurrence of injury were both taken into account, although this information was not listed for all drug monographs in all compendia. One example is spironolactone (AldactoneÒ), which was included in this study, but incomplete information exists to establish a cause-effect relationship. It was mentioned in Clinical Pharmacology that ‘‘spironolactone-induced fluctuations in serum electrolyte concentration can occur rapidly and precipitate hepatic encephalopathy in susceptible patients.’’ One group of susceptible patients were those who had pre-existing hepatic disease. It is not fully clear which other patient populations may be at a higher risk or if the risk is comparable in patients without baseline liver impairment. Similar to other studies, some potentially hepatotoxic drugs may have been inadvertently missed/excluded. Birth control pills are appropriate example. There are several cautionary statements about the use of combined oral contraceptive pills in women who have a history of liver impairment, hepatic cancer, or jaundice. Oral contraceptives have been associated with benign hepatic adenomas that may cause death, although this is very rare in the United States. Some recently approved new drugs might be also excluded in this study. For example, voriconazole (VfendÒ) was approved in May 2002 but was not found in the initial screen per literature search and the 1990–1997 Food and Drug Administration reports. This drug has since been linked to reversible hepatitis and jaundice. Despite increasing awareness of hepatotoxicity and the availability of less toxic alternatives, the absolute frequency of hepatic drug reactions has not decreased in the last decade.1–3,18,19,21 The present study could provide an overview of hepatotoxic drugs among the major drug compendium which can be used for future pharmacoepidemiological studies. The comparison of risk categories among 5 compendia provides useful information for physicians, pharmacists, nurses, and other clinical practitioners about potential hepatotoxic medications, and can also be used for patient consultation about medication use and hepatotoxic effects.

478

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

5. Conclusion Awareness of and monitoring for drug-induced hepatotoxicity is a critical issue among clinical practitioners. The rating scores of hepatotoxicity, emphasis on hepatotoxicity severity, and number of included hepatotoxic drugs differed among the studied drug compendia. USPDI, Facts & Comparisons, Clinical Pharmacology, and AHFS document large numbers of hepatotoxic drugs with detailed hepatotoxic information from both clinical trials and postmarketing epidemiologic evaluations. Comprehensive evaluations of hepatotoxic-related drugs provide critical information for health care practitioners and pharmacoepidemiologic researchers. References 1. Lee WM. Drug-induced hepatotoxicity. N Engl J Med. 2003;349:474–485. 2. Andrade RJ, Camargo R, Lucena MI, Gonzalez-Grande R. Causality assessment in drug induced hepatotoxicity. Expert Opin Drug Saf. 2004;3:329–344. 3. Rashid M, Goldin R, Wright M. Drugs and the liver. Hosp Med (London). 2004;65: 456–461. 4. Russo MW, Galanko JA, Shrestha R, Fried MW, Watkins P. Liver transplantation for acute liver failure from drug induced liver injury in the United States. Liver Transpl. 2004;10:1018–1023. 5. Shakya R, Rao BS, Shretha B. Incidence of hepatotoxicity due to antitubercular medicines and assessment of risk factors. Ann Pharmacother. 2004;38:1074–1079. 6. Teoh NC, Farrell GC. Hepatotoxicity associated with non-steroidal anti-inflammatory drugs. Clin Liver Dis. 2003;7:401–413. 7. Sulkowski MS. Hepatotoxicity associated with antiretroviral therapy containing HIV-1 protease inhibitors. Semin Liver Dis. 2003;23:183–194. 8. Spriet-Pourra C, Auriche M. Drug Withdrawal from Sales. 2nd ed. Richmond: PJB Publications; 1994. 9. Friis H, Andreasen PB. Drug-induced hepatic injury: an analysis of 1100 cases reported to the Danish Committee on Adverse Drug Reactions between 1978 and 1987. J Intern Med. 1992;232:133–138. 10. Almdal TP, Sorensen TI. Incidence of parenchymal liver disease in Denmark, 1981 to 1985: analysis of hospitalization registry data. The Danish Association for the Study of the Liver. Hepatology. 1991;13:650–655. 11. Galan MV, Potts JA, Silverman AL, Gordon SC. Hepatitis in a United States tertiary referral center. J Clin Gastroenterol. 2005;39:64–67. 12. De Abajo FJ, Montero D, Madurga M, Garcia Rodriguez LA. Acute and clinically relevant drug-induced liver injury: a population based case-control study. Br J Clin Pharmacol. 2004;58:71–80. 13. Andrade RJ, Lucena MI, Alonso A, et al. HLA class II genotype influences the type of liver injury in drug-induced idiosyncratic liver disease. Hepatology. 2004;39:1603–1612. 14. Kaplowitz N, Aw TY, Simon FR, Stolz A. Drug-induced Hepatotoxicity. Ann Intern Med. 1986;104:826–839. 15. Zimmerman HJ, Ishak KG. General aspects of drug-induced liver disease. Gastroenterol Clin North Am. 1995;24:739–757. 16. Ostapowicz G, Fontana RJ, Schiodt FV, et al. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med. 2002;137: 947–954.

J.J. Guo et al. / Research in Social and Administrative Pharmacy 1 (2005) 460–479

479

17. Lewis JH, Zimmerman HJ. Drug-induced liver disease. Med Clin North Am. 1989;73: 775–792. 18. Whitehead MW, Hainsworth I, Kingham JGC. The causes of obvious jaundice in South West Wales: perception versus reality. Gut. 2001;48:409–413. 19. Aithal GP, Day CP. The natural history of histologically proved drug induced liver disease. Gut. 1999;44:731–735. 20. Lumley C. Clinical toxicity: could it have been predicted? Pre-marketing experience. In: Lumley CE, Walder SR, eds. Animal Toxicity Studies. Their Relevance to Man. Lancaster: Quay Publishing; 1990:49–57. 21. Larrey D. Drug-induced liver disease. J Hepatol. 2000;32:77–88. 22. Carson JL, Strom BL, Duff A, et al. Acute liver disease associated with erythromycins, sulfonamides, and tetracyclines. Ann Intern Med. 1993;119:576–583. 23. Gardcia Rodriguez LA, Gutthann SP, Walker A, Lueck L. The role of non-steroidal anti-inflammatory drugs in acute liver injury. BMJ. 1992;305:865–868. 24. Lee WM. Acute liver failure. N Engl J Med. 1993;329:1862–1873. 25. Miwa LJ, Jones JK. Value of epidemiologic studies in determining the true incidence of adverse events. Arch Intern Med. 1997;157:2129–2136. 26. Aithal GP, Day CP. Hepatic adverse drug reactions. In: Mann RD, Andrews EB, eds. Pharmacovigilance. Hoboken, NJ: John Wiley & Sons, Ltd; 2002;459–474.