UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The Effect of acetaminophen on isoniazid metabolism Youssefi, Mohammed 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_youssefi_mohammed.pdf [ 5.04MB ]
Metadata
JSON: 831-1.0086189.json
JSON-LD: 831-1.0086189-ld.json
RDF/XML (Pretty): 831-1.0086189-rdf.xml
RDF/JSON: 831-1.0086189-rdf.json
Turtle: 831-1.0086189-turtle.txt
N-Triples: 831-1.0086189-rdf-ntriples.txt
Original Record: 831-1.0086189-source.json
Full Text
831-1.0086189-fulltext.txt
Citation
831-1.0086189.ris

Full Text

THE EFFECT OF ACETAMINOPHEN ON ISONIAZID METABOLISM by MOHAMMED YOUSSEFI B.Sc., The University of British Columbia, 1988  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES PHARMACOLOGY & THERAPEUTICS  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA © M. Youssefi, 1992  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of Pharmacology & Therapeutics The University of British Columbia Vancouver, Canada  Date 22 December, 1992  DE-6 (2/88)  ii  ABSTRACT Acetaminophen (APAP, paracetamol, N-acetyl-p-aminophenol), an analgesic and antipyretic drug, causes liver necrosis in overdose. Isoniazid (INH, isonicotinyl hydrazide), an antituberculous drug, also causes liver damage in some patients at therapeutic doses. The two medications are likely to be taken concurrently. Previously we reported that, depending on the condition, INH inhibits or induces the toxic pathway of APAP metabolism. In this study the influence of APAP on INH metabolism was investigated in ten healthy volunteers. INH, 300 mg, was ingested daily for 7 days. APAP, 500 mg, was ingested on the day before starting INH (day 0) and on day 7 of INH ingestion. INH and its major metabolites in urine (days 6 and 7) were analyzed by reversed-phase high-performance liquid chromatography. Two metabolites, isonicotinic acid (INA) and isonicotinylglycine (ING), were analyzed directly by using an isocratic mobile phase system; propionylisoniazid was used as an internal standard. INH and its hydrazine metabolites, hydrazine (Hz) and acetylhydrazine (AcHz), were derivatized with manisaldehyde and subsequently analyzed by using a gradient mobile phase system; 9-fluorenone was a standard. A different procedure was used to analyze acetylisoniazid (AcINH) and diacetylhydrazine (DiAcHz). The INH, AcHz, and Hz present in the samples were converted to hydrazones by reaction with pchlorobenzaldehyde. The hydrazones were then extracted with methylene chloride. The AcINH and DiAcHz remaining in the aqueous layer were then converted to INH and AcHz, respectively, by partial acid hydrolysis. The partial hydrolysis products were derivatized with m-anisaldehyde and then analyzed as above. The 2-tailed paired-sample t test suggested that concomitant APAP ingestion had no effect on the metabolism of INH, but this conclusion is far from certain. A number of explanations are given.  iii  TABLE OF CONTENTS CHAPTER^  PAGE  Abstract^  ii  Table of contents^  iii  List of Tables^  viii  List of Figures^  ix  List of Abbreviations^  x  Acknowledgements^  xiii  1  INTRODUCTION  1  1.1  GENERAL PRINCIPLES OF ISONIAZID  1  1.1.1  Chemistry of Isoniazid  1  1.1.2  Therapeutic Uses and Mechanisms of Action  2  1.1.2.1  Uses Against Mycobacterium Tuberculosis  2  1.1.2.1.1  Antimicrobial Spectrum and Mechanisms of Action  2  1.1.2.1.2  Treatment Regimens  3  1.1.2.2  Clinical Uses  4  1.1.3  Major Pathways of INH Metabolism  5  1.1.3.1  Acetylation of IN H  5  1.1.3.2  Hydrolysis Reactions and Glycine Conjugation  8  1.1.3.3  Acetylation of Hz and AcHz  8  1.1.3.4  Condensation With a-Ketoacids  8  1.1.3.5  Other Pathways of INH Metabolism  9  1.1.4  Microsomal Metabolism and Hepatotoxicity  10  1.1.4.1  Isoniazid  10  iv  TABLE OF CONTENTS (CONT'D) CHAPTER  PAGE  1.1.4.2  Acetylhydrazine  11  1.1.4.3  Hydrazine  12  1.1.5  Factors That Alter INH-induced Liver Toxicity  13  1.1.5.1  Effects of Age  13  1.1.5.2  Dose of I N H  13  1.1.5.3  Co-administration of Other Drugs  14  1.1.5.4  Cytochrome P450 Isozyme Composition  15  1.1.5.5  Species Differences  15  1.1.5.6  Acetylator Status  16  1.1.5.6.1  Epidemiological Studies  16  1.1.5.6.2  Metabolic Studies  17  1.1.6  Pharmacokinetics  19  1.1.6.1  Absorption and Bioavailability  19  1.1.6.2  Distribution  20  1.1.6.3  Renal Clearance  20  1.1.7  Adverse Effects  21  1.1.7.1  Hepatotoxicity  21  1.1.7.2  Neurotoxicity  21  1.1.7.2.1  Mechanism of CNS Toxicity  22  1.1.7.3  Treatment of INH Overdose  23  1.1.8  Future Developments  23  V  TABLE OF CONTENTS (CONT'D) PAGE  CHAPTER 1.2  GENERAL PRINCIPLES OF ACETAMINOPHEN  24  1.2.1  Chemistry of Acetaminophen  24  1.2.2  Therapeutic Uses  25  1.2.3  Mechanisms of Action  25  1.2.4  Biotransformation and Hepatotoxicity  26  1.2.4.1  Conjugation Reactions  26  1.2.4.2  Microsomal Metabolism  29  1.2.4.2.1  Nontoxic Route  29  1.2.4.2.2  Toxic Route  30  1.2.4.3  Mechanism of NAPO! Formation  31  1.2.4.4  Mechanisms of NAPO! Hepatotoxicity  31  1.2.5  Factors that Alter APAP Hepatotoxicity in Man  32  1.2.5.1  Effects of Age  32  1.2.5.2  Effects of Pregnancy  33  1.2.5.3  Interspecies Differences due to Enzyme Variations  33  1.2.5.4  Variability of P450 Isozyme Composition in Humans  33  1.2.5.5  Effects of Other Drugs  34  1.2.5.6  Effect of Diet  35  1.2.6  Pharmacokinetics  35  1.2.5.1  Absorption  35  1.2.5.2  Distribution  35  1.2.5.3  Excretion  36  1.2.7  Adverse Effects  36  1.2.8  Future Developments  37  vi  TABLE OF CONTENTS (CONT'D) CHAPTER^  PAGE  1.3 DRUG INTERACTIONS AND OBJECTIVE OF STUDY ^38 1.3.1^The Effect of INH on Drug Metabolism ^ 39 1.3.1.1^P450 Inhibition and Induction by INH ^ 39 1.3.1.2 The Effect of INH on APAP Metabolism ^41 1.3.2^The Effect of Other Drugs on INH Metabolism ^41 1.3.3^The Effect of APAP on Drug Metabolism ^41 1.3.4^Objective: The Effect of APAP on INH Metabolism ^42 2  METHODS^  43  2.1 MATERIALS^  43  2.1.1^Commercially Obtained Supplies^  43  2.1.2^Synthesis of Reference Compounds ^ 43 2.2 PROTOCOL^  45  2.2.1^Subjects^  45  2.2.2^Administration Regimen^  45  2.3 HPLC ANALYSIS^  45  2.3.1^Column^  46  2.3.2^Assays of INA and ING^  46  2.3.2.1^Preparation of Assay^  46  2.3.2.2 Apparatus and Chromatographic Conditions^46 2.3.3^Assays of Acetylhydrazine, INH, and Hydrazine^47 2.3.3.1^Preparation of Assay^  47  2.3.3.2 Apparatus and Chromatographic Conditions^49 2.3.4^Assays of Acetylisoniazid and Diacetylhydrazine ^50 2.3.4.1^Preparation of Assay^  50  VII  TABLE OF CONTENTS (CONT'D) CHAPTER^  3  PAGE  2.3.4.2^Apparatus and Chromatographic Conditions  52  2.4  CALCULATIONS  52  2.5  STATISTICAL ANALYSIS  52  RESULTS  53  3.1  RESULTS OF INA AND ING  53  3.2  RESULTS OF INH, AcHz, AND HZ  61  3.3  RESULTS OF ACETYLISONIAZID AND DIACETYLHYDRAZINE  67  3.4  SUMMARY OF RESULTS AND STATISTICAL ANALYSIS  70  3.4.1^Results of Paired Comparison Statistics  70  3.4.2^Graphical Analysis and Other Methods  77  3.4.2.1^Box plots  77  3.4.2.2^Scatter Plots and Variance Statistics  78  3.4.2.3^Hotelling Multivariate Statistics  79  3.4.2.4^Power of Test and Sample Size  80  4  DISCUSSION  91  5  REFERENCES  101  viii  LIST OF TABLES TABLE  PAGE  1  Summary of acetylator phenotypes, sex, and sample numbers of the subjects  56  2  Summary of INA and ING as % of total INH dose  60  3  Summary of INH, AcHz, and Hz as % of total INH dose  66  4  Summary of AcINH and DiAcHz as % of total INH dose  69  5  Summary of recovery of total isonicotinyl and total hydrazide compounds  73  6  The effect of acetaminophen on the hepatic conversion of isoniazid to its metabolites in 5 rapid and 5 slow acetylators  74  7  Summary of the statistical results  76  8  Minimum Percent Change Detectable  88  9  Power of Paired-Sample t test  89  10  Sample Size Required  90  ix  LIST OF FIGURES FIGURE  PAGE  1  The metabolic pathways of INH  6  2  The metabolic pathways of Hz and AcHz  7  3  The metabolic pathways of acetaminophen  27  4  The metabolism of NAPQI and 3-GSH-APAP,  28  5  Derivatization reactions of AcHz, INH, and Hz  48  6  HPLC chromatogram of INA and ING  57  7  The standard plot for INA  58  8  The standard plot for ING  59  9  HPLC chromatogram of AcHz, INH, and Hz  62  10  The standard plot for INH  63  11  The standard plot for AcHz  64  12  The standard plot for Hz  65  13  The standard plots for AcINH and DiAcHz  68  14  Scatter plots of distribution of differences  75  15  Box plots of total recoveries of isonicotinyl compounds in fast ans slow acetylators  82  16  Box plots of total recoveries of isonicotinyl compounds in men and women  83  17  Scatter plots of total recoveries of isonicotinyl compounds in fast and slow acetylators  84  18  Scatter plots of total recoveries of isonicotinyl compounds in men and women  85  19  Scatter plots of total recoveries of hydrazide compounds in fast and slow acetylators  86  20  Scatter plots of total recoveries of hydrazide compounds in men and women  87  LIST OF ABBREVIATIONS 3-Cys-APAP^3-cysteinylacetaminophen 3-Cys-Gly-APAP^3-cysteinylglycinylacetaminophen  3-GSH-APAP^3'-S-glutathionyl-acetaminophen 3-0CH3-APAP^3-methoxyacetaminophen 3-0H-APAP^3-hydroxyacetaminophen 3-SCH3-APAP^3-methylthioacetaminophen 9-Fl^ 9-fluorenone, volumetric internal standard AcHz^acetylhydrazine AcHz-KA^acetylketoglutaric hydrazone AcHz-PA^acetylpyruvic hydrazone AcINH^acetylisoniazid ACT^ Acyl-CoA transferase APAP^acetaminophen, paracetamol ATP^ adenosine triphosphate A.U.F.S.^detector absorbance units, full scale BNPP^bis-p-nitrophenyl phosphate, amidase inhibitor CNS^ Central Nervous System Conc^Concentration Cys^ cysteine day Day 6^INH only ingested by subjects Day 7^APAP + INH ingested by subjects DiAcHz^diacetylhydrazine DiINH^di-isonicotinylhydrazide  xi  LIST OF ABBREVIATIONS (CONT'D) F^  female subjects  9^  gram  GABA^gamma-amminobutyric acid GABA-T^GABA aminotransferase GAD^ glutamic acid decarboxylase Glu^ glutamic acid Gly^ glycine GSH^ glutathione, reduced form GSSG^glutathione disulfide HPLC^high-performance liquid chromatography hr^  hour(s)  Hz^  hydrazine  INA^ isonicotinic acid ING^ isonicotinylglycine INH^ isoniazid, isonicotinyl hydrazine INH-KA^isonicotinylketoglutaric hydrazone INH-PA^isonicotinylpyruvic hydrazone i.v.^  intravenous  kg^  kilogram  M^  male subjects  mg^ milligram MPO^ myeloperoxidase min^ minute (s) NAD^ nicotinamide adenine dinucloetide  xii  LIST OF ABBREVIATIONS (CONT'D) NAPO!^N-acetyl-p-benzoquinone imine NAPSQI^N-acetyl-p-benzosemiquinone imine NAT^ N-acetyltransferase, N-acetylase P450^ cytochrome P-450 P-Azine^pyruvate azine PABA^p-aminobenzoic acid PAH^ p-aminohippuric acid PAPS^phosphoadenosine phosphosulfate PB^  phenobarbital  PcINH^propionylisoniazid, internal standard PHR^ peak height ratio PIH^ pyridoxal isonicotinyl hydrazone PIP^ pyridoxal phosphate, biologically active form of pyridoxine p.o.^ oral administration R^  rapid/fast acetylators of INH  RMP^ rifampin, rifampicin s^  slow acetylators of INH  t;i^  apparent half-life  TB^  tuberculosis  THOPC^1,4,5,6-tetrahydro-6-oxo-3-pyridazine carboxylic acid UDPGA^uridine diphosphate glucuronic acid uv^  detector ultraviolet wavelength  Vd^  volume of distribution  Water-I^deionized water  ACKNOWLEDGEMENTS  I am very grateful to Dr. Richard A. Wall and Dr. James M. Wright for their support, encouragement, patience, and continuing advice. Without their support this work would not have been possible. Thanks are also due to Dr. M. J. Walker and Dr. M. Schulzer for consultation regarding statistical analysis.  Thanks are also extended to the members of the Department of Pharmacology & Therapeutics for getting me fascinated with the science of pharmacology and for their assistance in the completion of this degree. Special thanks are extended to Mrs. Maureen Murphy and Mr. Christian Caritey for providing technical assistance.  The secretarial staff of the department, JaneIle Swetnam, Margaret Wong, and Elaine Jan are thanked for their assistance.  The department is thanked for financial sponsorship in the form of teaching assistantships. The financial support of Challenge Awards is also greatly appreciated.  I also express my gratitude to the members of my Supervisory Committee (Dr. Richard A. Wall, Dr. James M. Wright, and Dr. Frank Abbott) for making the completion of this thesis possible at a very busy time of the year. All of their suggestions and criticisms are greatly appreciated.  1  1 INTRODUCTION 1.1 GENERAL PRINCIPLES OF ISONIAZID Tuberculosis (TB) is an infectious disease that has plagued humans since the dawn of history. Ancient civilizations made reference to this condition as early as 2500 B.C., yet it was not until 1882 that the etiological agent was defined as Mycobacterium tuberculosis. Although chemotherapy of TB has been practiced  for 2 millenia, only in 1939 did it became a clinical reality. Dapsone, a breakthrough in the chemotherapy of TB, was made an analog of the antibacterial sulfonamides (Lewis & Shepherd, 1970; Holdiness, 1985). About 1 billion people are infected with the tubercle bacillus, and 3 million deaths occur annually. Isoniazid (INH, isonicotinyl hydrazide), the most powerful and best tolerated tuberculostatic drug, was introduced clinically in 1952. Since then it has been used worldwide for the chemoprophylaxis and treatment of TB. The pharmacology of INH has been extensively studied. Recently the metabolism of INH has been studied in detail due to the discovery that the drug causes liver injury in some individuals at therapeutic doses (Mitchell et al., 1975a). In the last decade the metabolism of INH has been studied in this laboratory. This thesis is about a drug interaction: the effect of acetaminophen on INH metabolism.  1.1.1 CHEMISTRY OF ISONIAZID  INH (C6H7N30, F.W. 137.15, m.p. 170-174°C), colorless crystals or white crystalline powder, a weak base (pKa = 1.8, 3.5, and 10.8 at 20° C), is the hydrazide of isonicotinic acid. It is soluble in water and boiling alcohol, but not in organic solvents such as ether and benzene. INH forms complexes with divalent cations, and it reduces ammoniacal silver salt solutions to yield elemental silver. Rapid and thermodynamically favorable reactions of INH with aldehydes, ketones, a -ketoacids, and glucose yield hydrazones (Brewer, 1977).  2  INH was discovered by Meyer & MaIly (1912), but its tuberculostatic action was not recognized until its rediscovery in 1951. The quest for antituberculous drugs found nicotinamide, sulfathiadiazoles, thiosemicarbazones (e.g. thioacetazone; synthetic precursors in the preparation of sulfathiadiazole analogs), and cyclic N,N'-phthalylhydrazine to be active (Chorine, 1945; Domagk eta!., 1946, Behnisch eta!., 1950).  The strong antituberculous activity of INH was reported independently and simultaneously by 3 groups. Bernstein et al. (1952) made 5000 analogs of thioacetazone; INH was an intermediate. Following the thiosemicarbazone and nicotinamide leads, Fox (1952) studied many pyridine derivatives, including isonicotinaldehyde thiosemicarbazone. When methyl isonicotinate was treated with hydrazine, INH was obtained as the first of two intermediates. Offe et al. (1952) pursued the thioacetazone and phthalic hydrazide leads, modified the structure of the latter compounds and made the benzalhydrazone of INH; INH was an intermediate. The quest for more active and less toxic analogs of INH led to studies of structure-activity relationships. Only iproniazid matches INH in activity, but it has serious liver and CNS toxicity (Lewis & Shepherd, 1970).  1.1.2 THERAPEUTIC USES AND MECHANISMS OF ACTION 1.1.2.1 Uses Against Mycobacterium Tuberculosis 1.1.2.1.1 Antimicrobial Spectrum and Mechanism of Action  INH is unusually specific for all species of tubercle bacilli of mammals and birds, especially M. tuberculosis. Low concentrations (0.025-0.05 g/ml) inhibit mycobacteria, but, relatively high concentrations (600 g/m1) are required to inhibit other bacteria (Krishnamurti, 1975). INH does not affect the growth of fungi or protozoal parasites.  3  The precise molecular mechanism of action of INH is still controversial (lwainsky, 1988). INH is taken up by mycobacterial cells by active transport and then oxidized to isonicotinic acid (INA) by peroxidase (Krishnamurti, 1975). INA is then converted to 4-hydroxymethylpyridine, with an aldehyde as the intermediate, and then into isonicotinamide. Mycobacterium avium shows an enzyme activity designated as "hydrazidase" which converts INH to INA and hydrazine. The oxidation of INH to INA may generate reactive radicals (Shoeb eta!., 1985). Inhibition of nnycolic acid synthesis, peroxidase metabolism of INH and subsequent generation of oxygen free radicals, interference with NAD + and pyridoxal phosphate metabolism have all been proposed as possible mechanisms of action of INH. Zhang eta!. (1992) report that catalase-peroxidase may chemically convert INH to a biologically active form; catalase-peroxidase may play a role in the mode of INH action and nnycobacterial resistance. However, Quemard et al. (1991) claim just the opposite: intact INH is responsible for the tuberculostatic mechanism and peroxidase plays no important role. The loss of acid fastness of INH-treated cells suggested the hypothesis of inhibition of mycolic acid synthesis. Mycolic acids, unique to mycobacteria, are very long-chain (C24-C26) a- branched, 13- hydroxy fatty acids that cross-link the outer cell wall to the inner membrane. Apparently the target of INH is either synthesis of acetate metabolites used to elongate the fatty chains or some early steps of chain elongation (Quemard et al., 1991). Inhibition of cell wall building facilitates the attack on infecting mycobacteria by the host immune system.  1.1.2.1.2 Treatment Regimens  To combat TB, INH (5 mg/kg or 300 mg/d, p.o., i.m.) is taken for 6-12 months. The drug is taken daily, especially in the first 2 months of therapy. Patient compliance, drug toxicity, and the development of microbial resistance  4  present special therapeutic problems. The use of INH in conjunction with other drugs prevents resistance. Concomitant administration of pyridoxine (15-50 mg/d) minimizes adverse reactions in malnourished patients and those predisposed to neuropathy (Mandell & Sande, 1990).  1.1.2.2 Clinical Uses The therapeutic value of INH was examined in Huntington's Chorea (Perry et al., 1982), tremors associated with multiple sclerosis (Bozek et al., 1987;  Francis eta!., 1986), Parkinson's disease (Gershanik eta!., 1988), and rheumatiod arthritis. Only a small proportion of the multiple sclerosis and Huntington patients benefit from INH therapy. A mechanism has been proposed for the beneficial action of INH. The levels of gamma- aminobutyric acid (GABA), an inhibitory neurotransmitter, are markedly reduced in many brain regions of Huntington patients. Hydrazine (Hz), a metabolite of INH, inhibits GABA aminotransferase (GABA-T, an enzyme that degrades GABA) and increases GABA levels in the brain (Perry eta!., 1981, 1982, 1985). The mechanism of action might involve the carbonyl trapping of the co-factor, pyridoxal phosphate (PIP), by Hz. It has been documented, however, that the GABA-elevating effect of INH is not abolished by the coadministration of pyridoxine. Hz does not inhibit glutamic acid decarboxylase (GAD), an enzyme that also uses PIP as a co-factor and produces GABA (Perry eta!., 1981). A mechanism other than GABA elevation may underly the effect of INH, such as inhibition of monoamine oxidase (Francis et al., 1986). Another possibility is that clinical improvement in Huntington patients was masked or blocked by alterations produced in opposing neurotransmitter systems in which synthesis and/or degradation of the neurotransmitter is dependent on pyridoxal phosphate-mediated reactions (Manyam eta!., 1987).  5  1.1.3 MAJOR PATHWAYS OF INH METABOLISM  The biotransformation pathways of INH by the liver are shown in Figure 1. The main pathways involve acetylation, hydrolysis, and condensation reactions. Microsomal and peroxidase pathways are discussed later.  1.1.3.1 Acetylation of INH Arylamine N-acetyltransferase (NAT), a cytosolic enzyme, catalyzes the acetylation of a range of drugs and environmental chemicals, such as arylamines and hydrazines. There are two gene loci in humans encoding different isozymes of NAT. One gene locus (pnat, equivalent to NAT2) is multi-allelic and encodes polymorphic NAT. The other gene locus (mnat, equivalent to NATI), encodes monomorphic NAT and it does not vary amongst human individuals (Grant et al., 1990). The pnat locus products can confer either the fast (rapid) or the slow acetylator pharmacogenetic phenotype. NAT is inherited by simple autosomal Mendelian segregation of two alleles at a single gene locus. Persons homozygous (rr) for mutant alleles are slow acetylators, whereas those homozygous (RR) or heterozygous (Rr) for the wild-type allele are fast acetylators (Price-Evans, 1989). Slow acetylation may result from decreased synthesis of enzyme protein. Polymorphic NAT is expressed in the liver and small intestine, but monomorphic NAT is much more wide-spread in tissue distribution. INH was one of the first substrates to be recognized as being subject to polymorphic acetylation in humans. The acetylation of INH to acetylisoniazid (AcINH), a phase II reaction, is the main pathway in INH elimination in man (Hughes, 1953; Peters et al., 1965b; Boxenbaum & Riegelman, 1976). Slow and fast acetylators differ in their ability to acetylate INH; the rate of acetylation of INH is 4-5 times faster in the latter group. Consequently, there is a bi-modal distribution in the production of AcINH by fast and slow acetylators.  6  Figure 1. Metabolic pathways of isoniazid (INH). AcINH is the major metabolite in humans. Biotransformation enzymes include polymorphic N-acetyltransferase (NAT), amidase, acyl-CoA transferase (ACT), and cytochrome(s) P-450. INH, AcHz, and Hz also react chemically with a -ketoacids, such as pyruvate and a -oxoglutaric acid, to yield hydrazones or azines. The hypothetical toxication pathway(s) (---) generates free radicals that cause hepatotoxicity.  7  o^AcHz-KA U  CH3^CHs I^I C=N-N=C i i COOH COOH  COOH  Pyruvate Azine  CH3-C-NH-N=C-COOH 1 CH2-CH2-COON  HN — N  0 AcHz-PA I  THOPC  NH3  Urea  ....  Z NAT NH2 NH2 <^ Hz IP450 VIM  H2N-NH-OH ?‘ H2N-N=0  _  _  X.  0 II CH3-C-OH Acetic acid / CO2  H20 -  HN=NH  CH3-C-NH-N=C-COOH I CH3  ?^? ----> N2 •---  or X+ ....  NI, Cavalen/ binding ? to liver protein  0^iv,47 DiAcHz II CH3 C-NH-NH2 AcHz IP450  _  0 II CH3-C-NH-NH-OH 7^  0 II CH3-C-NH-N-0  IIMIN  -H20  0 II CH3-C-N=NH  0^0 II^II CH3- C+ , CH3-C• . CH2 = C = 0  _  2 Covalent binding Liver necrosis <--L--to liver protein  Figure 2. The metabolic pathways of Hz and AcHz. Both metabolites get acetylated  by NAT. Condensation reactions of Hz and AcHz with pyruvate yield pyruvate azine and AcHz-PA, respectively; reactions with a -oxoglutarate yield THOPC and AcHz-KA, respectively. Postulated microsomal pathways create hepatotoxins.  8  1.1.3.2 Hydrolysis Reactions and Glycine Conjugation In a phase I reaction in the cytosol, amidase catalyzes the hydrolysis of INH to isonicotinic acid (INA) and hydrazine (Hz). Amidase also catalyzes the hydrolysis of AcINH to INA and acetylhydrazine (AcHz); this pathway is the main route of formation of INA (Ellard eta!., 1972). Hz and AcHz are minor metabolites, however they are toxic. Via a phase II reaction, INA is conjugated with glycine by the nnitochondrial enzyme, acyl-CoA glycine transferase (ACT), to yield isonicotinylglycine (ING, isonicotinuric acid). Individuals differ in their ability to conjugate INA with glycine, but these differences are unrelated to acetylator status. INA was identified as a metabolite of INH as soon as INH came into clinical use (Kelley et a/., 1952). The first published clinical report of the metabolic fate of INH identified intact INH, INA, and ING in the urine (Cuthbertson et al., 1953).  1.1.3.3 Acetylation of Hz and AcHz In addition to acetylating INH, polymorphic NAT also acetylates Hz to AcHz (Fig. 2); AcHz is acetylated further to N,N'-diacetylhydrazine (DiAcHz)(Ellard & Gammon, 1976; Hein & Weber, 1982; Lauterburg et aL, 1985a). About 5% of Hz is acetylated to AcHz and about 10% is excreted unchanged (Wright & Timbrell, 1978). As expected, fast acetylators have a greater capacity to convert AcHz to DiAcHz than slow acetylators. A small amount of AcHz may be hydrolyzed to Hz and acetate by amidase (Timbrell eta!., 1980; Wright & Timbrell, 1978). Hz, AcHz, and DiAcHz were first identified as INH metabolites by McKennis eta!. (1959).  1.1.3.4 Condensation with a-Ketoacids INH undergoes nonenzymatic reactions with endogenous a -ketoacids such as pyruvic acid and a- oxoglutaric acid, to yield acid-labile hydrazones:  9  pyruvic acid isonicotinylhydrazone (INH-PA) and a- ketoglutaric acid isonicotinylhydrazone (INH-KA)(Zamboni & Defranceschi, 1954). The formation of INH-PA exceeds that of INH-KA by 2 or 3 times (Colvin, 1969). The extent of formation of hydrazones of INH depends on the amount of INH available for condensation (Peters et al., 1965b). In a similar fashion (Fig. 2), AcHz also reacts with the endogenous aketoacids to form pyruvic acid acetylhydrazone (AcHz-PA) and a -oxoglutaric acid acetylhydrazone (AcHz-KA)(Ellard & Gammon, 1976; Wright & Timbrell, 1978). A molecule of Hz reacts chemically with two molecules of pyruvic acid to yield a nontoxic metabolite, pyruvate azine (P-Azine). In vivo 15N-NMR studies in rat hepatocytes have revealed that Hz reacts with 2-oxoglutarate to form 2oxoglutaric acid hydrazone which immediately undergoes intramolecular cyclization to yield 1,4,5,6-tetrahydro-6-oxo-3-pyridazine carboxylic acid (THOPC). Hz is also metabolized to ammonia which is then converted to urea by the hepatic urea cycle (Nelson & Gordon, 1982; Preece eta!., 1991).  1.1.3.5 Other Pathways of INH Metabolism Di-isonicotinylhydrazide (DiINH) is apparently formed from the condensation of INH with INA. In mice and guinea pigs DiINH accounts for about 1-4% of the total urinary excretion of INH. Isonicotinamide and the glucose hydrazone of INH also appear to be INH metabolites (Peters et al., 1965a; Colvin, 1969; Brewer, 1977). INA could be converted to 4-hydroxymethylpyridine, with 4pyridyl aldehyde as the intermediate, and then into isonicotinamide. The aldehyde intermediate is difficult to detect because it is metabolized rapidly by cells. There is also the possibility for the direct conversion of INH to 4hydroxymethylpyridine (wall, personal communication).  10  1.1.4 MICROSOMAL METABOLISM AND HEPATOTOXICITY  INH, AcINH (Snodgrass eta!., 1974), AcHz (Nelson eta!., 1976; Timbrell et al., 1980; Wright & Timbrell, 1978), and Hz (Noda et al., 1983; Blair et al., 1985)  are believed to be responsible for causing liver injury. For a number of reasons, it is unlikely that AcINH itself is toxic: i) pretreatment of rats with a competitive amidase inhibitor bis-para-nitrophenyl phosphate (BNPP), a compound that inhibits the hydrolysis of AcINH to AcHz, prevents necrosis caused by AcINH (and INH), ii) but AcHz-induced necrosis can not be prevented by BNPP pretreatment, and iii) AcHz is a more potent hepatotoxin than AcINH (Mitchell et al., 1976; Timbrell et al., 1980). It is not known which hydrazine metabolite is  responsible for INH-induced liver injury in humans. In man, the metabolism, disposition, and toxicity of hydrazines are not well known (Preece et al., 1992). The oxidation of hydrazines apparently produces free radical intermediates (Noda et al., 1986) which may induce cellular toxicity by covalent binding to tissue macromolecules (Timbrell eta!., 1980). Free radicals from hydrazines can occur in a variety of oxidizing systems including cytochrome P-450/NADPH, FAD-dependent monooxygenases, peroxidases (Prough et al., 1981), prostaglandin synthase/arachidonic acid, cmhemoglobin, and neutrophil derived oxidants. The possibility exists that these chemicals cause liver injury by a mechanism independent of microsomal oxidation. Hydrazines are very good nucleophiles, and they may also covalently bind to cell components without metabolic activation.  1.1.4.1 Isoniazid In the last two decades liver injury and fatal intoxication cases caused by therapeutic doses of INH have been reported (Black et al., 1975). It is widely believed that 2 metabolites of INH, Hz and AcHz, are heptotoxins but not INH  11  itself. INH may be directly oxidized to INA and free radicals by cytochrome P450 2E1 (Wall, personal communication). In fact INH was oxidized directly to INA by myeloperoxidase present in activated leukocytes (Hofstra et al., 1992). The peroxidase-catalyzed direct oxidation of INH to INA may generate isoniazidyl and pyridinyl free radicals that may cause lipid peroxidation (Fig. 1)(Sinha, 1983; Shoeb et al., 1985). These reactive radicals could be responsible for hepatotoxicity.  1.1.4.2 Acetylhydrazine As shown in Figure 2, AcHz is eliminated by 3 processes (Timbre!' et al., 1977a): i) it is excreted intact in the urine and as hydrazones, ii) it is acetylated to DiAcHz, and iii) it is converted to reactive electophilic intermediates by the hepatic microsomal enzyme system. These radicals alkylate liver proteins and interact with thiols. The adducts produced with cysteine and glutathione have been characterized, revealing that the acetyl group is the alkylating species (Nelson et al., 1976a-b). The P450 oxidation also yields acetate, which is further metabolized  to CO2 (TimbreII eta!., 1980; Lauterburg eta!., 1985a). Basic studies led to the postulation that AcHz is responsible for INHinduced hepatotoxicity (Mitchell et al., 1975b; Mitchell et al., 1976; Nelson et al., 1976). The administration of AcHz (or AcINH) causes dose-related hepatic necrosis in rats pre-treated with PB. PB is an inducer of microsomal enzymes. Pretreatment of rat hepatocytes with inhibitors of cytochrome P450 such as CoCl2, carbon monoxide, SKF 525A, metyrapone, and 1-naphthyl-isothiocyanate inhibits the microsomal metabolism of AcHz (Albano & Tomasi, 1985). A number of workers have been unable to reproduce the rat model of AcHz liver injury (Wright et al., 1986). AcHz-induced hepatitis occurs only when rats are pretreated with PB (Mitchell eta!., 1976; Bahri eta!., 1982).  12  1.1.4.3 Hydrazine Hz was identified and isolated in the 1870s, and synthesized in 1907 (cited by Juchau & Horita, 1972). In nature Hz occurs in tobacco and mushrooms. Hz is a metabolite of INH and other drugs (Noda et al., 1978; Blair et al., 1985). Several drugs are derivatives of Hz. Hz is widely used in the industry in the preparation of blowing agents (used in the manufacture of plastics), boiler feed water (Hz is an oxygen scavenger in high temperature boilers), plant growth retardants, and spacecraft propellants (the combustion of Hz is very exothermic) (Preece et al., 1992). Basic studies indicate that Hz is a toxic compound that causes fatty liver, liver necrosis, disturbances of the CNS and red blood cell, and tumours in various organs (Moloney & Plough, 1983; Juchau & Horita, 1972; Colvin, 1969). Hz undergoes a variety of biotransformations including microsonnal oxidation, acetylation, and nonenzymatic Schiff's base formation with carbonyl compounds (Fig. 2). Studies in rabbits have shown that microsomal metabolism of Hz generates radical intermediates and nitrogen, probably from the diimide intermediate (HN =NH), hydroxyhydrazine ((H2NNHOH), and other hepatotoxic species (Noda eta!., 1983, 1985). Expired nitrogen is a major metabolite of Hz in mice in vivo (Nelson & Gordon, 1982). In mice 25-35% of the dose of Hz is expired as N2. Both oxyhemoglobin and liver P450 are capable of oxidizing Hz to N2, and diimide, a powerful diazene reducing agent (Nelson & Gordon, 1982; Springer et al., 1981). These reactive intermediates, like those of AcHz, are believed to bind covalently to hepatic macromolecules and cause hepatic injury. Despite the large number of studies conducted on the toxicity of Hz, the in vivo fate of Hz and its mechanism of hepatotoxicity are not well known. Recent pharmacokinetic studies suggest some correlation between levels of Hz and hepatotoxicity (Gent et al., 1992; Woo et al., 1992). Although the liver toxicity of  13  Hz may be due to its microsomal metabolism, the toxicity may be caused by intact Hz (TimbreII et al., 1982). Both in vivo and in vitro studies in rat hepatocytes suggest that ATP depletion may underlie the hepatotoxicity of Hz. Hz could deplete hepatic ATP by a number of possible mechanisms (Preece et al., 1990; Sannins et al., 1992).  1.1.5 FACTORS THAT ALTER INH-INDUCED LIVER TOXICITY  The risk of INH liver damage increases with increasing age, with increased dose, and with some co-administered drugs. Other factors that may influence INH-induced liver toxicity include genetic factors such as species differences, cytochrome P450 isozyme composition, and acetylator phenotype.  1.1.5.1 Effects of Age Age appears to be the most important factor in determining the risk of INHinduced hepatotoxicity. Age has no significant influence on acetylator phenotype or on the half-life of INH, but it is associated with a decrease in the apparent volume of distribution (Kergueris et al., 1986). Increasing age leads to a striking increase in the incidence of hepatic injury from INH (Black et al., 1975; Dickinson et al., 1981; Mitchell eta!., 1976): the incidence is rare in those patients less than  20 years old, 0.3% in those 20 to 34 years old, 1.2% in individuals 35 to 49, and 2.3% in individuals older than 50 years of age.  1.1.5.2 Dose of INH Wright & Wall (1985) provide clinical evidence that as the dose of INH increases, hepatotoxicity also increases. This is supported by a report that the incidence of hepatitis was appreciably lower with a 12 mg/kg dose as compared with 20 mg/kg in children with TB (Parthasarathy et al., 1986).  14  1.1.5.3 Co-administration of Other Drugs Upon co-administration with INH, carbamazepine induces hepatic microsomal enzymes and, subsequently, increases production of hepatotoxic INH metabolites (Wright et aL, 1982, 1984; Barbare et al., 1986). According to epidemiological studies, chronic alcohol consumption increases the risk of liver damage with INH (Dickinson et al., 1981). There are at least 2 possible mechanisms: i) induction of microsomal metabolism of INH by alcohol, and ii) possible additive hepatotoxicity. PB is a well-known inducer of the microsomal metabolism of INH. The effect of RMP on the microsomal metabolism of INH in humans is still unclear. RMP does not influence the metabolism of INH and its metabolites in man (TimbreII et al., 1985; Jenner & Ellard, 1989), however these authors did not measure the effect of AMP on Hz metabolism. In rabbits, RMP pretreatment increases the microsomal metabolism of Hz but not AcHz (Noda et al., 1984). Based on indirect measurements, Sarma et al. (1986) suggested that  RMP influences the metabolism of Hz. Pharmacokinetic studies of Hz in TB patients on multiple drug therapy suggest that AMP has no major effect on Hz formation (Gent et al., 1992). Concomitant use of INH and RMP increases liver toxicity, but the effect may be additive since RMP itself is known to cause liver damage (Steele eta!., 1991). Some compounds such as Cu2+ , zn2+, mn2+, Ni2+, sulfhydryl inhibitors, N-ethylmaleimide, and p-chloromercuribenzoate act as blocking agents of polymorphic NAT, but they are not used as drugs (Weber & King, 1981). Drugs sharing the acetylation pathway, such as procainamide, are able to interfere with the acetylation of INH (Ladero eta!., 1989). Prednisolone enhances acetylation of INH in slow acetylators (Sarma et al., 1980). However, such interactions are not clinically significant.  15  1.1.5.4 Cytochrome P450 Isozyme Composition Amongst humans, differences in the complement of P450 forms present in various tissues, especially in the liver, is a major factor responsible for interindividual variation in the metabolism and effects of drugs and toxic chemicals (Kalow, 1987). Individuals who have higher levels of the specific P450 isozymes that metabolize INH and/or its metabolites to toxic intermediates, should have a higher risk of INH-induced liver damage.  1.1.5.5 Species Differences Despite the fact that many papers appeared on INH hepatotoxicity, only a small number of researchers actually demonstrated liver necrosis in rats resulting from the actions of INH or one of its metabolites (Snodgrass et al., 1974; Erill et al., 1977; Bahri et al., 1981), and so far none of these groups have been able to  reproduce their original experiments. Covalent binding of liver macromolecules in rats has been demonstrated, but it was also shown that the pathological damage of the liver is not necessarily correlated to covalent acetyl binding to hepatic microsomal macromolecules (Wodward & TimbreII, 1984). Since the rat was not a good model for the study of INH-induced hepatotoxicity, rabbits were studied by Whitehouse et al. (1983) and our laboratory. Interestingly, NAT is polymorphically inherited in rabbits, like humans, but not in rats, mice, or guinea pigs. It appears that the levels of amidase in rabbits is higher than that in rat, and this allows liver damage to occur by INH alone ((Whitehouse et al., 1983). However, as Noda et al. (1983) have shown, and also demonstrated in our laboratory, Hz may be the cause of hepatotoxicity in rabbits not AcHz.  16  1.1.5.6 Acetylator Status Acetylator phenotype has been associated with differing susceptibility to disease. Slow acetylators are at a greater risk of predisposition to bladder cancer. Colonic cancer, in contrast, may be associated with the fast acetylator phenotype. Despite many epidemiological and metabolic studies and debates, the precise relationship, if any, between acetylator status and the incidence of INH-induced hepatotoxicity is unclear.  1.1.5.6.1 Epidemiological Studies  The results of epidemiological studies are conflicting. Evidence has been obtained that i) fast acetylators are at a greater risk of IN H-induced liver damage (Mitchell et al., 1975b, 1976); ii) slow acetylators are at a greater risk according to Dickinson eta!., 1981, Musch et a!., 1982, and Parthasarathy eta!., 1986; and iii) the risk is about the same in both groups (Gurumurthy eta!., 1984; Ellard eta!., 1978). The experimental conditions were not identical in the above studies. Mitchell did not have proper controls for age differences of his patients. It seems that the significance of the acetylator status as a predictor of susceptibility to liver damage is highly dependent on the index of liver damage chosen. If one chooses clinical jaundice as Gurumurthy et al. have done, "liver damage" is unrelated to acetylator status. On the other hand, if a biochemical index such as serum transaminase elevations are chosen (Dickinson et al., 1981; Musch et al., 1982), "liver damage" depends on acetylator status. Clearly, failure to define such terms adequately may lead to loss of valuable information if not invalidate any conclusions. In some of the studies multiple drugs were used for the treatment of TB; this may invalidate the conclusions. In other cases the therapeutic regimen was not the same as those of others.  17  1.1.5.6.2 Metabolic Studies Clinical studies with healthy subjects show that upon administration of INH, the levels of INH, AcHz, and Hz are higher in the plasma of slow acetylators than fast acetylators (Blair et al., 1985). Most metabolic studies suggest that slow acetylators are at a greater risk of INH-induced hepatotoxicity if AcHz is the causative agent. An initial report suggested that fast acetylators have a higher risk of INH-induced liver damage, presumably because they convert more INH to AcINH and therefore AcHz. It was confirmed that fast acetylator humans form higher amounts of AcHz, but they also detoxify more AcHz (by conversion to DiAcHz) than slow acetylators. After ingestion, about 23-28% of INH is excreted as DiAcHz by fast acetylators vs 5-8% by slow acetylators (Ellard & Gammon, 1976; TimbreII et aL, 1977a). Based on these measurements, TimbreII et al. estimated that the amount of AcHz that should undergo metabolic activation through the cytochrome P-450 pathway to hepatotoxicity in the fast acetylators (16.3%) should not be significantly different from the slow acetylators (18.2%). Based on results from the study of patients on triple therapy (INH, RMP, and ethambutol) for TB (Musch et al. (1982; Eichelbaum et al. (1982), fast acetylators (n=26) formed almost twice as much AcHz (40.1% of INH dose) as compared to slow acetylators (n=37; 25.8% of INH dose). In fast and slow acetylators about 70% (25.9% of INH dose) and 27% (4.9% of INH dose), respectively, of the AcHz formed is acetylated to DiAcHz. Renal excretion of AcHz and its hydrazones in fast and slow acetylators amounted to 2.3% and 3.5% of INH dose, respectively. In slow acetylators 68% of AcHz (17.5% of INH dose) is oxidatively metabolized, whereas in fast acetylators only 30% (12% of INH dose) is metabolized along this pathway. If AcHz is the compound responsible for hepatotoxicity, then this study suggests that the risk of hepatitis is higher in slow acetylators. Hz levels were not measured.  18  It was found that the formation of 14CO2 after administration of [14C)AcHz or [14q-AcINH in rats reflects the fraction of the dose that passes through the toxic pathway (TimbreII eta!., 1980). Building on these observations, a similar study was done in humans (Lauterburg et a!., 1985a). The cumulative exhalation of 14co2 increases as the rate of acetylation of INH decreases, such that slow acetylators generate more 14CO2 than fast acetylators. Thomas et a/. (1987) found that in rabbits, covalent binding of 14C to hepatic protein was inversely proportional to the rate of acetylation. This is an important finding if we asssume that covalent binding is a prerequisite for hepatotoxicity. These observations suggest that slow acetylators metabolize more INH via the toxic pathway and hence would probably be more susceptible to INH hepatitis, particularly at high doses of the drug. The acetylation of AcHz to DiAcHz is the most important detoxifying step (Wright & Timbrell, 1978). The balance between microsomal metabolism and acetylation of AcHz may be a key determinant of the toxicity of INH (Lauterburg et  al., 1985b). INH has been reported to alter the metabolism of AcHz. In vivo, INH inhibits the acetylation of AcHz to DiAcHz (Wright & Timbrell, 1978; Timbrell, 1979). The authors predicted correctly that normal therapeutic doses of INH in humans would cause inhibition of AcHz acetylation (Peretti et a!., 1987). Because the levels of INH (and AcHz and Hz) are higher in slow acetylators (Boxenbaum & Riegelman, 1976; Weber & Hein, 1979), acetylation of AcHz is probably more inhibited in slow as compared to fast acetylators. Lauterburg et a/. (1985b) did a pharmacokinetic study of INH metabolism in man. Since the half-life of elimination of AcHz is fivefold slower than that of INH, repeated doses of INH are expected to cause accumulation of AcHz. One might expect that, due to higher blood levels of INH, slow acetylators would have a greater risk of developing hepatotoxicity. When a large dose of INH such as 10 mg/kg is administered, saturation of  19  acetylation may occur, particularly in patients who are slow acetylators. This saturation of acetylation may be especially contributory to an accumulation of AcHz in slow acetylators by decreasing its clearance to DiAcHz (Lauterburg et al., 1985b). However, Bahri et al. (1981) showed histologically that INH, when coadministered with AcHz, decreased the liver injury of AcHz when the rats were pretreated with PB. This could be due to the inhibition of cytochrome P-450 mediated activation of AcHz by INH (Wright & TimbreII, 1978; TimbreII & Wright, 1979; TimbreII, 1979). INH inhibits liver microsomal enzymes in man (Kutt et al., 1970; Wright eta!., 1982). This may partly explain why several small doses of INH cause mild liver toxicity in PB-pretreated animals, but large single doses of INH do not (Mitchell et al., 1976). Timbrell & Wright (1979) found that INH inhibits the covalent binding of 14C-AcHz to rat microsomal proteins in vitro, but the concentration required is unlikely to be attained in vivo after therapeutic dose of the drug. Therefore, it is not clear if high levels of INH, as compared to low levels, enhance or inhibit the hepatotoxicity of AcHz in man.  1.1.6 PHARMACOKINETICS  1.1.6.1 Absorption and Bioavailability Since INH is a weak base that is effectively unionized at body pH and is both water and fat soluble and not significantly protein bound, it readily diffuses across lipid membranes. As a consequence, it is rapidly and completely absorbed after oral or parenteral dosage (Ellard & Gammon, 1976; Mitchell et a!., 1975b; Weber & Hein, 1979). Good estimates of possible loss due to first-pass metabolism are not available; therefore, the bioavailability of INH is not known (Wright, personal communication; Mandell & Sande, 1990).  20  1.1.6.2 Distribution INH and its 2 major metabolites, AcINH and INA, are distributed in total body water (Vd= 0.8 L/kg). The peak levels of INH attained in slow acetylators exceed those in rapid acetylators by about 25% (Ellard, 1984). Therapeutic doses of INH (5 ring/kg) cause peak plasma concentrations (Cm) of 3-5 A g/ml within 1-2 hours of oral ingestion (Weber & Hein, 1979). The therapeutic concentrations of INH in the plasma can range from 3 to 10 A g/ml. In the whole population, the half-life of INH generally averages 1 hr (fast acetylators) and 3 hrs (slow acetylators). Hepatic insufficiency prolongs the half-life of INH.  1.1.6.3 Renal Clearance Intact INH, its hydrazones, AcINH, INA, ING, intact AcHz, its hydrazones, DiAcHz, intact Hz and its derivatives, get excreted in the urine. About 85-98% of the total dose of INH is excreted in the urine. About 0.5-10% of the total dose is excreted in the bile (lwainsky, 1988). From 75 to 95% of a dose of INH is excreted in the urine within 24 hours, mostly as metabolites. INH undergoes glomerular filtration and tubular reabsorption. Elimination of INH depends mainly on its metabolic biotransformation. The clearance of INH is dependent to only a small degree on the status of renal function, however patients who are slow acetylators may accumulate toxic concentrations if their renal function is impaired (Mandell & Sande, 1990). AcINH also undergoes glomerular filtration. INA and ING are actively secreted. The apparent first-order rate constants for the excretion of some of the metabolites of INH have been determined (Ellard & Gammon, 1976).  21  1.1.7 ADVERSE EFFECTS  Hepatitis and peripheral neuritis are the most serious and the most common side effects, respectively, of INH use. Other adverse effects include hypersensitivity/allergic reactions, idiosyncratic reactions, lupus erythrematosus, etc. (Mandell & Sande, 1990).  1.1.7.1 Hepatotoxicity Liver toxicity secondary to INH therapy was first reported one year after the drug's remarkable effectiveness against TB was recognized. Most INH-induced hepatic damage occurs 4 to 8 weeks after the start of therapy. Chronic administration of therapeutic doses of INH causes subclinical hepatic dysfunction (which is mild and reversible) in 10 to 20% of patients (Mitchell et a!., 1975a, 1976). The overall incidence of overt INH-induced hepatitis is ca. 1-2% (Black et al., 1975; Mitchell et al., 1976; Dickinson et a!., 1981; Gurumurthy et a!., 1984); up  to 10% of these patients may die from hepatic failure.  1.1.7.2 Neurotoxicity If pyridoxine (vitamin B6) is not given concurrently, peripheral neuritis (especially in slow acetylators) is the most common reaction to INH and occurs in about 2% of patients receiving 5 mg/kg of the drug daily. Higher doses may result in peripheral and central nervous system (CNS) toxicity in 10-20% of patients. These toxic reactions include peripheral neuritis, insomnia, restlessness, urinary retention, lactic acidosis, muscle twitching, convulsions (when the dose is over 17 mg/kg), psychotic episodes, seizures and coma (Mandell & Sande, 1990; Ellard, 1984). Acute neurological toxicity, associated with overdose, may be fatal if not recognized and treated promptly.  22  1.1.7.2.1 Mechanism of CNS Toxicity  The adverse effects produced by INH intoxication may be due to the lowering of the effective tissue and serum levels of pyridoxine through the formation of two types of pyridoxal isonicotinyl hydrazones (PIH's) which are rapidly excreted through the kidney. The PIH's also cause competitive inhibition of pyridoxine kinase, an activating enzyme, which converts pyridoxine to the physiologically active pyridoxal phosphate (PIP). As mentioned before, therapeutic doses of INH inhibit GABA-T and consequently raise GABA levels in the brain (Perry et al., 1982). GABA is an inhibitory neurotransmitter in the CNS. GAD and GABA-T are involved in the synthesis and degradation of GABA, respectively. PIP is an important cofactor in the metabolism of multiple neurotransmitters. Many enzymes require PIP as a cofactor, e.g., apotryptophanase (which degrades tryptophan). PIP is important in the metabolism of GABA, because it is a cofactor for both GAD and GABA-T. It has been speculated that overdoses of INH causes inhibition of GAD. INH-induced seizures may be produced by an overwhelming stimulation of the CNS, apparently caused by i) depletion of GABA with subsequent heightened CNS excitability, and ii) inhibition of nnonoamine oxidase which leads to increased sympathetic activity (Cash & Zawada, 1991). Most of these problems can be effectively prevented by daily administration of supplemental pyridoxine. Another adverse effect, lactic acidosis, caused by INH overdose is postulated to be induced by seizure activity, which results in an anerobic state. Alternatively, INH may also interfere with nicotinamide-adenine dinucleotide (NAD) that is necessary for the conversion of lactate to pyruvate in the Kreb's cycle.  23  1.1.7.3 Treatment of INH Overdose Lethality from INH overdose has occurred from an oral dose of as little as 5 grams, although patients have ingested as much as 30 grams and survived (Leibowitz et al., 1989). Toxic effects have been generally associated with plasma concentrations greater than 20 A g/ml. The standard procedure is used for the treatment of INH overdose: lavage, emesis (but not in conscious patients), followed by activated charcoal. Vasopressors are occasionally required. Forced diuresis is not recommended. Intravenous pyridoxine in a dose equivalent to that of INH ingested aborts and prevents seizures. Hemodialysis is used in severe cases (Cash & Zawada, 1991). Charcoal hemoperfusion was shown to be much more effective than standard hemodialysis (Leibowitz et aL, 1989). Severe hepatotoxicity does not appear to be a consequence of acute overdose of INH (Cash & Zawada, 1991).  1.1.8 FUTURE DEVELOPMENTS  At present INH is usually administered for 6-12 months for preventive therapy. However, this approach has many deficiencies. These include the expense of treating and monitoring patients for such a long time, noncompliance with preventive therapy of long duration, and the occurrence of drug toxicity. Because hepatotoxicity is a major limitation to the use of INH preventive therapy, research efforts should be encouraged that i) contribute to the understanding of the mechanisms of INH-induced hepatitis, ii) contribute to the understanding of the mechanism by which INH acts, which will in turn facilitate the development of congeners that retain antimicrobial potency while eliminating sites of hepatotoxic potential, iii) investigate the cytoprotective effects of drugs that may block or interfere with the the toxic effects of INH, and iv) identify more specific risk factors, other than age, for the development of INH-hepatitis (CDC, 1989).  24  1.2 GENERAL PRINCIPLES OF ACETAMINOPHEN Acetaminophen^(paracetamol,^4'-hydroxyacetanilide,^N-acetyl-paminophenol, APAP), a widely used analgesic and antipyretic medication that has weak anti-inflammatory action, has been available in most countries since the 1950s. It has become a common household analgesic because it lacks the occasional side effects of aspirin, is well tolerated, and is available without prescription (Penna & Buchanan, 1991). The drug is remarkably safe when used in therapeutic doses, but when taken in overdose, it causes acute centrilobular hepatic necrosis in man (Black, 1984; Hinson, 1980; Mitchell & Jollow, 1975; Prescott et al., 1983) and animals (Mitchell et al., 1973a). The pharmacological properties of APAP have been reviewed by Clissold (1986).  1.2.1 CHEMISTRY OF ACETAMINOPHEN  APAP (C8H9NO2, F.W. = 151.16, m.p. 169-172°C, white crystals or crystalline powder), a moderately water- and lipid-soluble weak organic acid (pKa =9.5), is largely un-ionized over the physiological range of pH. APAP is synthesized by the electrolytic reduction of nitrobenzene to p-aminophenol, followed by acetylation with acetic anhydride. In industry APAP is used as a photographic chemical and in the manufacture of azo dyes (Merck Index, 1989; Forrest et a!., 1982; Fairbrother, 1974). Acetanilide, accidentally discovered to have antipyretic action about 1886, was too toxic. In the search for less toxic analogs, p-aminophenol, phenacetin, and APAP were tested. APAP was first synthesized by Morse in 1878 by reduction of p-nitrophenol with tin in glacial acetic acid (cited by Fairbrother, 1974). The p-aminophenol produced by the reducing action of the tin was acetylated by acetic acid. APAP was first tested clinically as an antipyreticanalgesic drug by von Mering in 1893 (cited by Prescott, 1983). APAP became  25  popular in 1949, when it was found that i) it is the active metabolite of acetanilide and phenacetin, and ii) it is less toxic than the other 2 drugs (Insel, 1990). The antipyretic activity of APAP and other 4-aminophenol derivatives resides in the aminobenzene structure. Some APAP analogs, such as N-methylAPAP and 2,6-dimethyl-APAP, do not cause liver injury but unfortunately they lack analgesic activity. Methylation of APAP at position 2 or 3 of the benzene ring retains the analgesic activity but has hepatotoxicity as well. Positional isomers of APAP, 2- and 3-hydroxyacetanilide, retain some pharmacological activity and they are not as hepatotoxic as APAP. Some 3,5-dialkyl analogs of APAP are more potent than APAP itself but their hepatotoxic effects are controversial. None is as effective as APAP in terms of analgesia (Vermeulen et a!., 1992).  1.2.2 THERAPEUTIC USES  APAP is useful in mild to moderate pain, especially when patients are allergic to aspirin or when salicylates are poorly tolerated. Effective analgesicantipyretic levels of APAP are 10-20 1.4g/m1 (Insel, 1990). The conventional oral dose of APAP is 325-1000 mg; the total daily dose should not exceed 4 g.  1.2.3 MECHANISMS OF ACTION  The rapid, reversible, and non-competitive inhibition of cyclooxygenase by APAP involves antioxidant or free radical trapping properties (Lands, 1985). This effect is significant since it reduces the hydroperoxides, which are believed to have an essential role in the generation of prostaglandins by cyclooxygenase. In areas of inflammation, hydroperoxides generated by activated leukocytes negate the effect of APAP; this may explain the weak anti-inflammatory action of APAP. However, in conditions where leukocye infiltration is not a major factor, such as pain and fever, APAP is an effective medication.  26  1.2.4 BIOTRANSFORMATION AND HEPATOTOXICITY  The biotransformation pathways of APAP by the liver are shown in Figs. 3 and 4. The major pathways of APAP metabolism are conjugation reactions and P450 oxidative pathways that lead to both catechol and glutathione metabolites (Andrews et al., 1976; Mitchell et al., 1974; Slattery et a/., 1987). Large doses of APAP saturate the conjugation pathways and a higher amount is cleared through the microsomal oxidative pathways (Davis et al., 1976; Mitchell et al., 1974; Slattery eta!., 1987). Small amounts of deacetylated (4-aminophenol) and hydroxylated metabolites have also been found. The P450 metabolism of 4-aminophenol is apparently responsible for the nephrotoxicity of APAP in rats (Klos et al., 1992). The metabolites of APAP are therapeutically inert.  1.2.4.1 Conjugation Reactions The conjugations of APAP and some of its metabolites with UDPglucuronic acid (UDPGA) and phosphoadenosine phosphosulfate (PAPS) lead to glucuronide and sulfate conjugates, respectively (Figs. 3 & 4). APAP glucuronide and APAP sulfate together account for about 70-80% of therapeutic doses of APAP in humans (Prescott, 1983). The proportion of APAP eliminated as the sulfate vs. the glucuronide conjugate depends on the dose. The ratio of glucuronide to sulfate conjugates in human urine is usually between 1.8 and 2.3. Sulfation is a high-affinity, low capacity pathway whereas glucuronidation is a lowaffinity, high-capacity pathway (Hinson et al., 1981). At high doses of APAP the conjugation pathways become saturated and greater proportions of the drug are metabolized by P450.  27  Figure 3. The biotransformation pathways of acetaminophen (APAP) are catalyzed  by microsomal (*) and cytosolic (**) enzymes. Glucuronidation and sulfation are the major pathways at therapeutic doses of APAP; these pathways get saturated at overdoses of APAP, and more toxic metabolite, NAPO', gets formed. When most of the glutathione (GSH) gets depleted, NAPQI may cause hepatotoxicity by covalent attachment to macromolecules or by other mechanisms. +6 indicates that the meta position of NAPQI has a partial positive charge; this part of the molecule is attached to GSH or to macromolecules.  28  Figure 4. The fate of NAPQI, the toxic metabolite of APAP. NAPQI generated from  therapeutic doses of APAP is detoxified easily by glutathione (GSH); the GSH-derived conjugates account for 5-10% of the dose of APAP. Overdoses of APAP lead to depletion of GSH; the excess NAPQI then causes liver damage by a number of postulated mechanisms. Glutamyl transpeptidase and cysteinyl-glycinase are found on the extracellular face of the plasma membrane. +s indicates that the meta position of NAPQI has a partial positive charge; this part of the molecule is attached to GSH or to macromolecules.  29  1.2.4.2 Microsomal Metabolism Liver microsomes contain several P450 isoforms that are involved with the oxidation of APAP, including the ethanol-inducible form from rabbits, rats, and humans (P450 2E1; Morgan  eta!., 1983; Ryan et al., 1986), the 13 -naphtoflavone-  inducible form in rats and humans (P450 1A2; Harvison  eta!., 1988a; Raucy eta!.,  1989), the pregnenolone-16a-carbonitrile (PCN)-inducible form in rats (P450 3A1), and other isoforms such as 2C11 and 1A1 in rats (Lee et al., 1991). P450 1A2 metabolizes APAP mainly to a relatively nontoxic catechol metabolite, but it can also oxidize APAP to NAPQI. P450 2E1, P450 3A1, and myeloperoxidase (MPO/H202) oxidize APAP to NAPQI. Purified human liver P450 2E1 and its ortholog in rabbits (P450 3A) are relatively good catalysts of APAP oxidation to NAPQI, but only at high APAP concentrations (Nelson, 1990).  1.2.4.2.1 Nontoxic Route The oxidation of APAP by cytochrome P450 1A2 generates the catechol metabolite, 3-hydroxyacetaminophen (3-0H-APAP), which is then converted to 3methoxyacetaminophen (3-0CH3-APAP). After therapeutic doses in man, these aromatic oxidation products account for approximately 5 to 10% of the dose (Slattery et al., 1987; Raucy  et al., 1989). The catechols are interesting  metabolites because, like the thioether metabolites, their formation clearances decrease with increasing dose (Slattery  et al., 1987), and they are found in large  amounts in urines of overdosed patients (Andrews eta!., 1976). Some reactive quinones (benzoquinone, the quinone of the APAP catechol metabolite, and the quinone imine of the methylated catechol metabolite) are formed to a lesser extent, but their role in hepatotoxicity is not clear (Nelson, 1990).  30  1.2.4.2.2 Toxic Route As mentioned above, APAP is oxidized to NAPQI by a number of P450 isoforms. APAP causes hepatic necrosis in both animals and man when ingested in high enough doses. The exact sequence of events that ultimately leads to hepatic necrosis has not been elucidated. In man the oxidation of APAP by hepatic cytochrome P450 2E1 and 3A4 (and perhaps others) generates N-acetylp-benzoquinone imine (NAPQI), an electrophilic hepatotoxic metabolite (Raucy et al., 1989; Lee et al., 1991). The detoxification of NAPQI occurs by at least 2 mechanisms; NAPO! oxidizes reduced glutathione (GSH) to glutathione disulfide (GSSG), with the concomitant regeneration of APAP, and NAPQI reacts with GSH to form 3'-S-glutathionyl-acetaminophen (3-GSH-APAP). Events leading to cell death are initiated by the depletion of GSH and protein-bound thiols or by the binding of NAPO! to other constituents of the liver cell once GSH stores are depleted (Albano et al., 1985a; Dahlin eta!., 1984). The degradation of 3-GSH-APAP in the liver, gut and the kidney leads to the formation of other metabolites. About 5 to 10% of the dose of APAP is excreted as glutathione-derived conjugates (Fig. 4): 3-cysteinyl (3-Cys-APAP), 3mercapturate (APAP-3-mercapturate) and 3-methylthio (3-SCH3-APAP) conjugates. These thioethers are detoxication products of the reactive, putative toxic metabolite of APAP, NAPQI. Therefore they are indicators of flux through the toxic pathways. Cytochrome P450 can function as a peroxidase as well as an oxygenase (Nelson et al., 1981). There is a direct two-electron oxidation of APAP to NAPO! by cytochrome P450 and MPO or, alternatively, a one-electron oxidation to Nacetyl-p-benzosemiquinone imine (NAPSQI) by peroxidase, prostaglandin H synthase (PGS) or cytochrome P450 (Dahlin et al., 1984; Harvison et al., 1988b; van de Straat et al., 1988). The relationship of NAPSQI to APAP-induced  31  hepatotoxicity is unclear. NAPSQI can also react with molecular oxygen to form superoxide or other active oxygen species responsible for lipid peroxidation.  1.2.4.3 Mechanism of NAPQI Formation The mechanism of NAPQI formation from APAP is still unclear. Because different isozymes of cytochromes P450 form NAPQI and a catechol metabolite (3-0H-APAP) at different rates (Harvison et al., 1988a), two different schemes were proposed. Initial reactions are abstraction of hydrogen radicals by a ferryloxy form of P450 from either the phenolic oxygen (in the case of 3-0H-APAP) or amide nitrogen (in the case of NAPQI). Radical recombination reactions would then occur so rapidly that the radicals do not diffuse from the enzyme active site. Such a mechanism is in accord with what is known about the formation of radicals by cytochromes P450 and their subsequent fate. It is proposed that the recombination products essentially undergo a dehydration reaction to yield the observed two electron oxidation products, 3-0H-APAP and NAPQI.  1.2.4.4 Mechanism of NAPQI Hepatotoxicity When the GSH levels fall to 20-30% of normal, the excess metabolite (NAPQI) is free to produce cell damage (Mitchell et al., 1973a, b, 1974) Hepatic glutathione plays a vital role in protecting against APAP hepatotoxicity. Upon depletion of GSH, NAPQI causes liver injury by covalent binding and/or oxidative stress (thiol oxidation or lipid peroxidation). There is not enough evidence for the role of lipid peroxidation. Covalent binding appears to be the most damaging event. The resulting hepatotoxicity has often been correlated with the extent of the covalent binding of APAP to proteins (Mitchell eta!., 1973a; Jollow eta!., 1973; Potter et al., 1973; Hinson, 1980; Albano et al., 1985). The protein-bound  32  residues in vitro and in vivo were identified as 3-cysteinyl thioether conjugates (Nelson, 1990). Studies with positional isomers of APAP show that covalent binding may not be necessarily related to liver damage. For example, some positional isomers bind to the same extent to hepatic proteins, and also form GSH conjugates, but do not cause liver damage. APAP, however, damages the mitochondria much more than its isomers; this is consistent with the known oxidative properties of APAP (Dahlin eta!., 1984; Albano eta!., 1985). Thiols in the mitochondria may be an important target in the pathogenesis of APAP hepatotoxicity (Nelson et al., 1991). The mitochondrial proteins that are arylated have not been characterized yet. APAP, primarily through its reactive metabolite NAPQI, depletes cellular thiols, which in turn disrupts Ca2+ and ATP homeostasis. These changes can lead to increases in cell Ca2+ levels, mitochondrial Ca2+ cycling, and activation of proteases and endonucleases, which may be involved in the propagation of liver damage (Nelson, 1990). One can not rule out the possibility that both covalent binding and oxidant stress caused by APAP are important factors in liver cell damage, and their relative importance may be dependent on the redox state of the cell. It is to be expected that the mechanism of APAP liver injury is quite complex and multifactorial.  1.2.5 FACTORS THAT ALTER APAP HEPATOTOXICITY IN MAN  1.2.5.1 Effects of Age APAP overdose in young children has generally been associated with a much lower incidence of hepatotoxicity than the incidence in adults (Penna & Buchanan (1991). Limited studies in humans show an increased clearance of  33  APAP in young subjects, and this may be the result of increased clearance via sulfation.  1.2.5.2 Effects of Pregnancy Clearance of APAP in pregnant women through pathways of glucuronidation and oxidation (as measured by thioether conjugates) is increased significantly (about 2 times) in the third trimester (Miners eta!., 1986). While these effects were not seen in pregnant mice, GSH was depleted at a faster rate and the animals were more susceptible to APAP-induced liver injury (Larrey et al., 1986). However, an increased incidence of liver injury caused by APAP has not been documented in pregnant women.  1.2.5.3 lnterspecies Differences Due to Enzyme Variations Genetic effects on the oxidation of APAP to its hepatotoxic metabolite or metabolites have been observed in laboratory animals and man. Species differences in hepatotoxicity and metabolism are quite dramatic (Gregus et al., 1988). Hamsters are the species most susceptible to toxicity, followed by mice. Rats, rabbits, and guinea pigs are relatively resistant to APAP-induced liver injury. Mice and hamsters have high levels of the alcohol-inducible P450 2E in the liver, but rats have low levels. The more susceptible species clear a larger fraction of the dose through NAPQI. The susceptibility of humans is somewhere in the range between the mouse and the rat.  1.2.5.4 Variability of P450 Isozyme Composition in Humans The oxidation of APAP to NAPOI by human liver samples varied over a tenfold range, which suggests that P450 isozyme composition may play an important role in man when large doses are taken; not all subjects taking an  34  overdose of APAP are equally susceptible to its hepatotoxic effects (Prescott, 1983; Seddon et al., 1987; Parkinson & Hurwitz, 1991). In fact it was recently shown that the interindividual variation in the amount of hepatic P450 2E1 is considerable (Ingelman-Sundberg et al., 1990). The data of Ingelman-Sundberg et al. might indicate the presence of polymorphically distributed variant P450 2E1  genes which may be linked to the incidence of alcohol-induced liver damage.  1.2.5.5 Effects of Other Drugs The risk of APAP-induced hepatotoxicity is increased in alcoholics, presumably because of the induction of the toxic pathway of APAP metabolism (Seeff et al., 1986). Anticonvulsants such as phenytoin, PB, carbamazepine, and  primidone may increase the risk of APAP-induced hepatotoxicity (Minton et al., 1988; Smith et al., 1986). There is evidence to support the hypothesis that P450 2E1 may contribute to the enhanced susceptibility of chronic alcoholics to APAP hepatotoxicity (Lieber, 1988). However, when given acutely, ethanol decreases the metabolism of APAP to toxic metabolites and prevents APAP hepatotoxicity in animals (Sato & Lieber, 1981) and possibly in man (Critchley eta!., 1983). The activity of liver microsomal enzymes is also important since this may determine the rate of NAPO' formation. The hepatotoxicity of APAP is increased in most species by pretreatment with PB, pregnenalone 16 -carbonitrile, 3methylcholanthrene, and reduced by inhibitors such as SKF 525A, piperonyl butoxide, ketoconazole, cimetidine, and cobaltous chloride (Mitchell et al., 1973a; Jollow et al., 1973). Cimetidine has been shown to protect mice and man from APAP toxicity (Rolband et al., 1991). Aniline, a P450 2E1 substrate, competitively inhibits the activation of APAP by INH in mice (Seddon eta!., 1987).  35  1.2.5.6 Effect of Diet Prolonged fasting, ingestion of protein-deficient diets, or diethyl maleate administration, have all been shown to reduce the liver levels of GSH and increase the hepatotoxicity of APAP. Dietary supplementation of fish-oil protects mice against APAP hepatotoxicity in vivo (Speck & Lauterburg, 1991).  1.2.6 PHARMACOKINETICS 1.2.5.1 Absorption Although APAP is rapidly and almost completely absorbed from the gut, it is incompletely available to the systemic circulation after oral administration, a variable proportion being lost through first-pass metabolism (Perucca & Richens, 1979). The bioavailability appears to depend upon the amount administered, decreasing from 90% after 1-2 g to 68% after 0.5 g (Forrest eta!., 1982). APAP absorption depends on the rate of gastric emptying. The mean halftime for absorption from the small intestine is about 7 minutes, whereas gastric emptying half-times are about 3 times longer. In man, APAP absorption appears to be negligible from the stomach but very rapid from the small intestine. Absorption is by passive transport with first-order kinetics (Forrest et a/., 1982).  1.2.5.2 Distribution The concentration in plasma reaches a peak in 30 to 60 minutes, and the t1/2 in plasma is about 2-4 hours after therapeutic doses. Therapeutic levels in plasma range from 10 to 20  II g/ml.  With toxic levels of APAP or liver disease, the  half-life may be increased. APAP is relatively uniformly distributed throughout most body fluids. Binding of APAP to plasma proteins is variable; about 20 to 50% is bound at plasma concentrations associated with overdosage. The volume of distribution is about 0.95 I/kg.  36  APAP distributes throughout most tissues and fluids. After i.v. doses, plasma APAP concentration-time curves are multiexponential and the relatively short half-times for the initial disposition phase (t½ce equal to 3 to 19 minutes) indicate rapid tissue distribution (Forrest et a!., 1982).  1.2.5.3 Excretion After therapeutic doses, 90 to 100% of the drug may be recovered in the urine within the first day (Insel, 1990), with about 4, 55, 30, 4 and 4% appearing as unchanged APAP, and its glucuronide, sulfate, cysteine and mercapturic acid conjugates, respectively. The minor metabolites are also detected in the urine. The total body clearance of APAP is 5.0 ± 1.4 ml/min/kg (Insel, 1990). Since APAP is a moderately lipid-soluble weak organic acid (pKa =9.5) which is not extensively bound to plasma proteins, it is likely to undergo considerable glomerular filtration with subsequent passive tubular reabsorption. In healthy adults the renal clearance of APAP at normal urine flow rates is about 12 ml/min (range 5 to 20 ml/min). The highly polar glucuronide and sulfate conjugates appear to be both filtered at the glomerulus and actively secreted by the tubules since their renal clearances are approximately 130 and 170 ml/min, respectively. The clearance of these metabolites is independent of changes in urine flow and pH (Forrest et al., 1982). These two metabolites accumulate in subjects with impaired renal function.  1.2.7 ADVERSE EFFECTS  The side effects of APAP may include hepatotoxicity, nephrotoxicity, skin rash, other allergic reactions, neutropenia, thrombocytopenia, and hypoglycemic coma in overdose. In adults, hepatotoxicity may occur after ingestion of a single  37  dose of 10 to 15 g (150 to 250 mg/kg) of APAP; doses of 20 to 25 g or more are potentially fatal. Clinical signs of liver damage become apparent in 2-4 days. Perhaps 10% of poisoned patients who do not receive specific treatment develop severe liver damage; of these, 10 to 20% eventually die of liver failure. Severe liver damage occurs in 90% of patients with plasma concentrations of APAP greater than 300 A g/ml at 4 hours or 45 A g/ml at 15 hours after the ingestion of the drug. Minimal hepatic damage can be anticipated when the drug concentration is less than 120 A g/mL at 4 hours or 30 A g/ml at 12 hours after ingestion (Prescott, 1983; Insel, 1990). Early diagnosis is vital in the treatment of APAP overdoses. Decontamination in overdose cases include gastric lavage and emesis. Supportive therapy may be necessary. N-acetylcysteine, cysteamine, or methionine given within 10 hours can prevent impeding liver toxicity. Nacetylcysteine is now the preferred therapy. These antidotes are believed to act primarily by stimulating GSH synthesis and hence facilitate GSH conjugation of NAPQI.  1.2.8 FUTURE DEVELOPMENTS  Future studies need to focus on the time course of the hepatotoxic events caused by APAP to determine what kind of intervention may be appropriate at particular time periods in the treatment of APAP poisoning. Ideally, a nonhepatotoxic APAP analogue could be developed. It is unlikely that APAP, which is a relatively safe and inexpensive analgesic and antipyretic, will be replaced unless an analog can be developed with greater anti-inflammatory activity that retains the non-gastrointestinal irritant properties of APAP (Nelson, 1990).  38  1.3 DRUG INTERACTIONS AND OBJECT OF STUDY A significant change in the magnitude and/or duration of action of one drug (index drug) by concomitant or prior use of a second drug is defined as a drug interaction. The interaction can be desirable, adverse, or insignificant. Adverse drug interactions account for about 0.2% of all in-hospital adverse drug reactions (Wright, 1992). Drug interactions involve either pharmacokinetic or pharmacodynamic mechanisms. Pharmacokinetic interactions include many distinct processes such as drug absorption, distribution (including protein binding), hepatic metabolism, and renal excretion. Among these, modifications of hepatic metabolism appear to constitute the major source of drug interactions. The consequences of metabolic drug interactions are often difficult to predict, especially if a number of different pathways are involved. Hepatic drug interactions are usually the result of inhibition or induction of liver enzymes. Most inhibitory interactions that are of any clinical significance involve the microsomal oxidative enzymes. These interactions are not always predictable because there are many different mechanisms of inhibition and also because some inhibitors can also cause induction under certain conditions (e.g. chronic ethanol, isoniazid, etc.). Some drugs that inhibit P450s include cimetidine, disulfiram, acute ethanol, erythromycin, INH, ketoconazole, oral contraceptives, etc. Also, there are important interindividual variations in both P450-dependent catalytic activities and amounts of enzymes in human liver microsomes (Guengerich, 1989). The selective induction of a subset of the reactions catalyzed by the P450 monooxygenases suggested that distinct forms of P450 are induced by specific inducers. Studies carried out with laboratory animals have revealed the existence of a multigenic superfamily (Nebert & Gonzalez, 1987) and made possible the  39  classification of the inducers into 5 distinct groups, according to the form(s) of P450 specifically induced: i) polycyclic aromatic hydrocarbons (1A), ii) PB (2B), iii) chronic ethanol, INH, acetone, benzene, ether, imidazole, pyridine (2E), iv) glucocorticoids such as PCN and macrolide antibiotics, and AMP (3A), and v) clofibrate (4A). Induction is sometimes associated with de novo synthesis of the P450(s). Other possible mechanisms, such as protein stabilization, are discussed by Zand et al. (1992). P450 enzyme induction by xenobiotics can be species-specific as  well as tissue-specific (Alvares & Pratt, 1990). Some drugs can induce their own metabolism (autoinduction). Some enzyme inductions make it necessary to for the clinician to increase the dose of the index drug appropriately. Some enzyme inductions are harmful. For example, carbamazepine induces the microsomal metabolism of INH and increases formation of hepatotoxic INH metabolites (Wright, 1992; Wright eta!., 1982, 1984; Barbare eta!., 1986).  1.3.1 THE EFFECT OF INH ON DRUG METABOLISM  INH is involved in a number of complex drug interactions. INH can inhibit or induce the metabolism of many other drugs.  1.3.1.1 P450 Inhibition and Induction by INH INH inhibits the microsomal metabolism of carbamazepine (Wright et al., 1982), phenytoin (Kutt eta!., 1970), APAP (Epstein et al., 1991; Zand et al., 1992), theophylline (Samigun et al., 1990), benzodiazepines (Ochs et al., 1981, 1983), and valproic acid (Dockweiler eta!., 1987; Jonville eta!., 1991). Higher than usual doses of INH have been reported to inhibit the metabolism of warfarin (Baciewicz & Self, 1985). The interactions of INH with carbamazepine and phenytoin are clinically important, the former is actually a double-drug interaction. The  40  interactions involving carbamazepine, phenytoin, theophylline, valproic acid, and warfarin are more important in slow acetylators of INH. The microsomal enzyme inhibitory action of INH may be due to a metabolic intermediate of INH that forms a complex with the enzyme. This action of INH affects many P450 enzymes and appears to be nonspecific (Muakkassah et al., 1981; Moloney eta!., 1984; Zand eta!., 1992). The major routes of elimination of phenytoin and carbamazepine are 4-hydroxylation (P450 209; Veronese et al., 1991) and 10,11-epoxidation (P450 3A3/4; Kerr eta!., 1991), respectively. P450 2C9 also 7-hydroxylates S-warfarin; this is the major route of inactivation of this anticoagulant in humans (Gonzalez, 1992). Theophylline is oxidized by P450 1A1/2 (Sarkar et al., 1992). INH seems to inhibit both microsomal oxidative pathways of APAP, namely P450 1A2 and 2E1 (Zand et al., 1992). In humans the microsomal metabolism of APAP by P450 3A4 is also likely to be inhibited by INH. Triazolam and diazepam are metabolized by P450 3A4 (Kronbach et al., 1989) and the P450 2B/3A family (Inaba et al., 1988), respectively, in humans. These studies suggest that INH does not selectively inhibit one specific P450 form; it inhibits many isoforms. Studies in experimental rats have shown that INH induces hepatic microsomal P450 2E1 (Ryan et al., 1986) and potentiates the hepatotoxicity of APAP (Burk eta!., 1990) and halothane (Rice eta!., 1987) in rats and halothane in guinea pigs (Lind et al., 1991). INH also induces the same P450 isozyme in rabbits and it induces enflurane defluorination in humans (Mazze et al., 1982; Hoffman et al., 1989). In these studies, INH administration was stopped before the administration of the other drugs.  41  1.3.1.2 The effect of INH on APAP Metabolism We previously reported that INH inhibits the oxidative metabolism of APAP (Epstein et al., 1991). The effect of INH on the metabolism of APAP is actually quite complex (Zand et al., 1992). Chlorozoxazone, a drug metabolized primarily by P450 2E1 (Peter et al., 1991), was used to test the inductive effect of INH on this isozyme. Apparently when INH is present in the body it stabilizes P450 2E1 and inhibits its catalytic activity, but when INH is eliminated from the body there is induction of this P450 isozyme. Consequently, depending on the condition, INH inhibits or induces the toxic pathway of APAP metabolism. In fact, cases of APAP hepatotoxicity have been reported following the ingestion of long-term INH and high doses of APAP (Moulding et al., 1991; Murphy et al., 1990). The levels of INH and its metabolites were not measured in any of these cases. Murphy et al. (1990) report an individual who had taken about 11.5 g of APAP 12 hr after a dose of INH. The hepatotoxicity was most likely due to induction of microsomal P450 2E1 metabolism of APAP by isoniazid. It is unlikely that the high dose of APAP per se could have caused the liver damage; in one of the cases the level of APAP in blood was reported, and that level is generally not associated wih hepatotoxicity.  1.3.2 THE EFFECT OF OTHER DRUGS ON INH METABOLISM  The effects of carbamazepine, ethanol, and procainamide on the metabolism of INH are discussed in section 1.1.5.3.  1.3.3 THE EFFECT OF APAP ON DRUG METABOLISM  At doses of 1.6 g/d or more, APAP may inhibit the microsomal oxidative pathway of warfarin, but this has not been substantiated (Boeijinga et al., 1982; Bartle & Blakely, 1991). APAP competitively inhibits the sulfation of ethinyl  42  estradiol, an oral contraceptive (Rogers et al., 1987). APAP might also competitively inhibit the glucuronidation of chloramphenicol (Spika & Aranda, 1987; Stein eta!., 1989).  1.3.4 OBJECTIVE: THE EFFECT OF APAP ON INH METABOLISM  The interaction between APAP and INH is of clinical interest for a number of reasons: i) each medication can cause liver injury under conditions which can be altered when other drugs are taken concomitantly (Mitchell et al., 1976; Black, 1984; Nelson, 1990), ii) the 2 drugs are likely to be taken concurrently because TB patients ingest INH chronically for many months and APAP is a major nonprescription analgesic-antipyretic medication, and iii) after many years of decline, TB cases have been rising since 1985. The excess cases were most likely caused by the increased risk of TB in persons infected with the human immunodificiency virus (HIV). Widespread undernutrition and the spread of HIV are likely to worsen TB morbidity in the future. The treatment of mycobacterial infections will become an even more important and challenging problem (CDC, 1989; Starke, 1989). If these trends continue, combinations of INH/APAP could become even more common in the near future. The effect of APAP on the risk of INH-induced hepatotoxicity and INH metabolism is currently unknown (Epstein eta!., 1991). The aim of this study was to examine the effect of a single therapeutic oral dose of APAP on INH metabolism.  43  2 METHODS 2.1 MATERIALS 2.1.1 COMMERICALLY OBTAINED SUPPLIES  INA was obtained from Eastman Organic Chemicals (Rochester, NY, USA), AcHz (95% pure), m-anisaldehyde (3-methoxybenzaldehyde), and pchlorobenzaldehyde were purchased from Aldrich Chemical Co. (Milwaukee, WI, USA). Hydrazinium sulfate (Hz-sulfate) and 9-fluorenone were obtained from Fisher Chemical Co. (Fair Lawn, NJ, USA). N,N'-diacetylhydrazine was purchased from ICN Pharmaceuticals (Plainview, NY, USA). All the hygroscopic compounds were dried in the dessicator before making up standard solutions. INH, dichloromethane, HPLC grade acetonitrile and methanol, propan-/-ol, formic acid, acetic acid anhydride, diethyl ether and all acids and bases were obtained from British Drug House (BDH) Chemicals Canada Ltd. The mobile phases were filtered through a 0.2A m cellulose acetate filter (Micron Separations, Inc., Westboro, MA, USA) prior to use. The urine samples were filtered with 0.45A m filters prior to use (Millipore Corp., Bedford, MA, USA). Sulfamethazine required for the determination of acetylator phenotype was obtained from Mattheison, Coleman and Bell Co.  2.1.2 SYNTHESIS OF REFERENCE COMPOUNDS  AcINH was semi-synthesized by the method of von Sassen et al. (1985), wherein INH was reacted with a four-fold excess of acetic acid anhydride at room temperature for 1.5 hr with continuous stirring. After recrystallization from methanol-diethyl ether (1:4), colorless and needle-like crystals were obtained with a melting point of 164° C (uncorrected). The literature value of the melting point is 162-163° C; von Sassen et al. obtained 157-160° C. HPLC analysis of the AcINH  44  showed that contamination by INH, AcHz, or Hz was below the level of detection (ca. 0.2-0.4% at the detector sensitivity used). PropionylINH (Pcl NH, 1-isonicotiny1-2-propionyl-hydrazine, C9Fl1 1 N302) was synthesized by a modification of the method of von Sassen et al. (1985) for propionylhydrazine. PcINH was synthesized by reacting a four-fold excess of the methyl ester of propionic acid with INH. The reaction was allowed to proceed for 12 hr at room temperature with continuous stirring. The excess methyl propionate was removed by vacuum distillation. PcINH was recrystallized in dichloromethane/n-hexane (2:1, v/v). The melting point of the final product was 133-138°C, which is close to the literature values of 130.5-131.5°C (von Sassen eta!., 1985).  ING was prepared by a slight modification (Gardner et a/. (1954) of the original method (Rohrlich, 1951) for the preparation of nicotinylglycine. Fox & Field (1943) had previously reported a similar method for the preparation of nicotinylglycine. These methods used the acid azide method. One normal NaOH was used in place of 0.1N NaOH, and 1M HCI was subsequently used to achieve neutralization. The product was recrystallized a few times from water/ethanol (8:2, v/v) mixture, melting point 220° C (uncorrected) with decomposition. Various melting points have been reported in the literature: 230-232° C (Gardner et al. 1954), 224-225° C (corrected) with decomposition (Boxenbaum et al., 1974), and 256-258° C (El-Naggar et al., 1985). HPLC analysis showed that no INA was detected (or very little less than 1%). We were not able to synthesize ING by the acid chloride method of El-Naggar et al. (1985). However, Kubo et al. (1990) reported that they synthesized ING from isonicotinyl chloride and glycine.  45  2.2 PROTOCOL 2.2.1 SUBJECTS  The subjects, acetylator phenotyping, 24 hr urinary creatinine excretion, and clinical laboratory measurements are described in the first part of this study (Epstein eta!., 1991).  2.2.2 ADMINISTRATION REGIMEN  Isoniazid (300 mg p.o.) daily was administered for 7 days. Acetaminophen (500 mg p.o.) was given on the day prior to starting isoniazid (day 0) and on day 7 of isoniazid administration. Twenty-four hour urine samples were collected over 3 g of ascorbic acid and maintained at 4° C, on days 0, 6, and 7. Sample volumes were measured and small amounts were stored at -70° C until analysis.  2.3 HPLC ANALYSIS Various analytical procedures have been used for quantitative analysis of INH and its metabolites in blood and urine. These include colorimetric, fluorometric, microbiological, radioimmunological, mass fragmentography and more recently gas chromatographic and liquid chromatographic assays. INH exhibits a number of color reactions with many colorimetric reagents. Colorimetric techniques are relatively insensitive and cumbersome. Fluorometric techniques are more sensitive than colorimetric methods; both methods have been used to measure INH only. Microbiologic procedures, although highly sensitive, are too time consuming to perform, and this method does not differentiate INH from its hydrazones. Gas chromatographic (Timbrell et al., 1977b; Noda et al., 1978) or highperformance liquid chromatographic (Lauterburg et al., 1981; Holdiness, 1982; Blair et al., 1985; von Sassen et al., 1985; Svensson eta!., 1985; Jenner & Ellard,  46  1987) procedures alone, or in combination with mass spectrophotometry, enable INH and its metabolites to be measured simultaneously. The advantages and disadvantages of various methods have been reviewed (Holdiness, 1985).  2.3.1 COLUMN  INH and all of its metabolites were separated by HPLC using a reversedphase column (125 X 4.6mm, i.d.) packed with Spherisorb 5/./ m ODS 2 (Phase Separations Ltd., Norwalk, CT, USA).  2.3.2 ASSAYS OF INA AND ING  2.3.2.1 Preparation of Assay PcINH was the internal standard. The urine samples were filtered with Millipore filters, portions were aliquoted, PcINH was added, and the samples were diluted 40-fold with deionized water (water-l).  2.3.2.2 Apparatus and Chromatographic Conditions Sample volumes of 40 Al were injected per analysis. Ion-pair HPLC was used to analyze the two metabolites. The prepared samples were eluted isocratically at ambient temperature (approximately 22° C) with a mobile phase that consisted of 22% v/v acetonitrile, 3% v/v glacial acetic acid, 1 mM each SDS, triethylamine, and H3PO4 made up with deionized water (water-I). The flow rate was 0.8 ml/min (ca. 1200 psi). The mobile phase was filtered and then degassed in situ by helium. The two metabolites were analysed using a Waters Model M-  6000A chromatography pump (Waters Associates, Milford, MA, USA) equipped with a CE 212 variable wavelength UV detector (Cecil Instruments, Cambridge, UK), connected to an Apple Ile microcomputer; a chromatography program was  47  used (Interactive Microware, Inc., State College, PA, USA) to measure peak heights and/or areas. Quantification was based on uv absorbance at 267 nm.  2.3.3 ASSAYS OF ACETYLHYDRAZINE, INH, AND HYDRAZINE  2.3.3.1 Preparation of Assay The derivatization technique was previously developed in our laboratory (Wall, personal communication). The derivatizing reagent was made up as follows: 250 ilL of 3-methoxybenzaldehyde, 5 ml of 10mM 9-fluorenone solution (made up in propan-l-ol), and 3.75 ml of formic acid (minimum assay 88%) were put in a volumetric flask and made up to 50 ml with propan-/-ol. The final solution consisted of 41.1mM m-anisaldehyde (i.e. excess concentration), 1mM 9fluorenone, and 7.5% formic acid (v/v). The solution was transfered to a dark bottle and kept in the dark when not in use. The urine samples containing INH and its metabolites were diluted appropriately and then filtered. The urine solution and the derivatizing reagent were mixed in the ratio of 4:1 (e.g., 200 p I of INH solution was put in a plastic vial, and 50 th I of the reagent was added, and the vial was inverted a few times). The reaction was allowed to occur for 2 hr at room temperature in the dark. The acid present in the derivatizing reagent set the urine pH to 3 so that the a -ketoacid hydrazones of AcHz and INH, and the a -ketoacid azines of Hz could be readily hydrolyzed. The AcHz, Hz, and INH released in these reactions then reacted with 3-methoxpenzaldehyde to form N-acetyl-3-methoxybenzalhydrazone, 3methoxybenzaldiazine,^and^isonicotiny1-3-methoxybenzaldhydrazone, respectively (Fig. 5).  48  Figure 5. The derivatization of AcHz, INH, and Hz with m-anisaldehyde (3-  methoxybenzaldehyde). The reactions were allowed to occur at pH 3 for 2 hours at room temperature. Because m-anisaldehyde and 9-fluorenone are both photosensitive, the reaction vessels were kept in the dark.  49  Incubation of samples at pH 3 for 2 hr led to the hydrolysis of virtually all the a -ketoacids, and probably most of the azines. The azines of Hz are a little more stable than the hydrazones of INH and AcHz. A vast excess quantity of manisaldehyde made it very unlikely for the free a -ketoacids to react with INH and AcHz again (Wall, personal communication). The unreactive compound, 9-fluorenone, served as an internal check on volumetric accuracy. The derivatized products were analyzed with the HPLC system described below. Sample volumes of 80  gI  were injected per analysis.  2.3.3.2 Apparatus and Chromatographic Conditions AcHz, INH, and Hz were separated by using a 125mm-length column. The derivatized products were analysed using a Spectra-Physics SP8000B liquid chromatograph (Santa Clara, CA, USA) fitted with a Spectra-Physics SP8400 uv/vis detector set at 300 nm. A gradient system was set up. The oven temperature was 48° C, and the flow rate was 1 ml/min. Mobile phase A consisted of 20% MeCN/ 10% Me0H solution (unbuffered, for cleaning column). Mobile phase B consisted of 20% MeCN/ 10% Me0H, and 5mM Sodium acetate. The stock solution of sodium acetate was 1M (pH =5). Mobile phase C consisted of 40% MeCN/ 40% Me0H, and 5mM sodium acetate. The solutions were made up with water-I, and the buffered solutions were filtered before use and degassed in situ. The HPLC gradient for the 125mm-column was: Time (min.) %B %C 100 0 0 40 5 60 100 0 10 100 0 16.66 100 0 17.5  50  2.3.4 ASSAYS OF ACETYLISONIAZID AND DIACETYLHYDRAZINE  2.3.4.1 Preparation of Assay Some of the urine samples were diluted with blank urine. The samples were then filtered as usual. The procedure used to prepare samples involved 3 steps. First, the AcHz and INH that were present in the urine samples were derivatized with p-chlorobenzaldehyde to form hydrazones which were then extracted with CH2Cl2. The procedure is now discussed in more detail. The derivatizing reagent was a propan- /-ol solution containing 41.09mM pchlorobenzaldehyde and 7.5% (v/v) formic acid. One ml of this derivatizing reagent and 1 ml of sample urine were added to a 15mL eppendorf glass test tube (reaction vessel), which was capped and inverted a few times. The reaction was allowed to occur at room temperature in the dark for 2 hr. The derivatization reaction allowed AcHz and INH (present in the reaction vessel) to react with p-chlorobenzaldehyde to yield hydrazones. These hydrazone products were then eliminated by extraction with CH2Cl2. CH2Cl2 (4 ml) was added to the solution 3 times and each time most of the organic portion (bottom layer) was carefully removed with Pasteur capillary pipets. The residual CH2Cl2 was blown off with a stream of nitrogen gas. AcINH and DiAcHz remained in the aqueous layer. Second, the AcINH and DiAcHz remaining in the aqueous layer were then converted to INH and AcHz, respectively, by partial acid hydrolysis. In partial acid hydrolysis a fraction of the total amount of AcINH and DiAcHz gets hydrolyzed. In complete acid hydrolysis, all of the AcINH and DiAcHz molecules would get hydrolyzed. Partial acid hydrolysis was done by adding 100 A I of concentrated HCI (12.1M) to the reaction vessel (1.1M HCI in the reaction mixture), the reaction  51  vessels were capped, and then placed in a hot water bath. The reaction vessels were heated at 80°C for 10 min. The reactions are shown below:  Heating for a longer period of time (or a higher HCI concentration, or a higher water bath temperature) would hydrolyze all of the AcINH and DiAcHz. However, it was not necessary to do this since proper standard solutions were also prepared at the same time. At the end of the 10 minutes, the hydrolysis reaction was stopped by putting all the reaction vessels in cold water for 1 min. Then 500 p.I of 1.04M sodium citrate was added to each reaction vessel to partly neutralize the solution and bring the pH up to about 2.9. The solutions were filtered once more, and 200 p. 1 portions were aliquoted into 4 or 5 plastic vials and frozen at -70° C for analysis  the next day. Third, each plastic vial was taken out of the freezer and the contents were allowed to thaw. Then 50 41 of the derivatizing reagent (41.1 mM m-anisaldehyde  52  and 9-fluorenone in propan-/-ol) was added and the reaction was allowed to proceed for 2 hr. These prepared samples were subsequently analyzed as above (i.e., same as the procedure for INH and AcHz from here on).  2.3.4.2 Apparatus and Chromatographic Conditions The same SP8000B system was used to analyze hydrolysis products of AcINH and DiAcHz.  2.4 CALCULATIONS Peak height ratios were calculated by dividing the peak heights of INA and ING by that of the internal standard (PcINH), and the peak heights of AcHz, INH, Hz, AcINH, and DiAcHz by that of 9-fluorenone (unreactive internal standard). Calibration curves were constructed by plotting peak height ratios versus the concentration of the respective compound (II M). The concentrations of INH and its metabolites were calculated from the standard calibration plot using the slope (y =mx) obtained by a linear regression analysis. Quantification of all the compounds was achieved by comparison of their peak heights with that of their respective internal standards.  2.5 STATISTICAL ANALYSIS The parametric two-tailed paired-sample t test was used (Zar, 1984). For each drug or metabolite analysed, the differences were calculated within each pair of measurement (days 6 and 7) for each subject. P < 0.05 was accepted as an indication of significance. Single-sample Hotelling T2 statistic was also used (Zar, 1984). In addition, boxplots and scatter plots were prepared (Chambers et al., 1983).  53  3 RESULTS The ten healthy volunteers were between the ages of 23 and 56 years. Five subjects were slow acetylators (2 women and 3 men) and the other five were rapid acetylators (3 women and 2 men). All subjects completed the study without adverse effects. The data from all subjects were analyzed and comparison of INH metabolism was made on days 6 (INH) and 7 (INH + APAP). Table 1 shows the subject acetylator phenotypes, sex, and urine sample numbers.  3.1 RESULTS OF INA AND ING Fig. 6 shows sample chromatograms of INA and ING. The retention times (tR) and absorbance units full-scale (AUFS) are also shown. AcINH was also analyzed by this method. However something present in the urine (most likely ascorbic acid, as explained below) interfered with the separation of AcINH. Therefore, AcINH was analyzed by another method. All urine samples (blanks, etc.) were diluted 40-fold with deionized water before analysis. Each diluted sample contained 10 pg/m1 (51 /.1 M) PcINH (internal standard). The standards (for calibration plots) were made up in deionized water. Figure 6 (A) shows the chromatogram of a blank urine that does not  contain any INH, APAP, or ascorbic acid (sample 148M); PcINH (peak 4) has a retention time (tR) of 12.1 min. Figure 6 (B) displays a chromatogram of a urine sample from day 0 (APAP  was taken, urine was collected over ascorbic acid). The arrow will be discussed later. Peak 4 is that of PcINH. Peak 5 (tR =9.84 min) belongs to either ascorbic acid, or a metabolite of APAP; this peak does not appear in chromatogram A (where the subject took no APAP, and the blank urine contained no ascorbic acid).  54  Figure 6 (C) shows the chromatogram of a standard solution containing 20 AM each of INA (peak 1, tR =3.89 min), AcINH (peak 2, tR =9.40 min), and ING (peak 3, tR = 10.8 min). The solution also contained 10 A g/ml of PcINH (peak 4, tR =12.1 min). All the four peaks were resolved well. Figure 6 (D) displays the chromatogram of a urine sample from day 6 (219M). The subject had taken INH only, and the urine was collected over ascorbic acid. INA has a low retention time because its peak shows up just after the peaks of unretained compounds. The Arrow (chromatogram B) points to the position where the peak of INA would appear. Therefore, the peaks of unretained compounds do not interfere too much with the peak of INA. Peak 5 ( please see chromatogram B) appears again; the reason is because the AcINH peak has a shoulder. Also, peak 5 starts at about 9.70 min and ends about 9.98 min. The peak of AcINH normally ends at about 9.93 min. Therefore the 2 peaks interfere with each other. For some reason, whenever there is such an interference, the tR of AcINH goes down. The the tR of AcINH is 9.40 min in chromatogram C (standard solution) and its about 8.9 min in chromatogram D (INH ingested, sample collected over ascorbic acid). Therefore, it seems likely that peak 5 belongs to ascorbic acid. The tR's of INA, ING, and PcINH are very similar to those in chromatogram C. Figure 6 (E) shows the chromatogram of a urine sample from day 7  (246M). The subject had taken INH + APAP, and the urine was collected over ascorbic acid. As expected, there is interference between peaks 5 (ascorbic acid?) and 2 (AcINH). The tR's of INA, ING, and Pc1NH are very similar to those in chromatograms C and D. For the above reasons, the results of AcINH were considered unreliable. Furthermore, the standard plot of AcINH was not linear above 20 A M; above this  55  concentration, the plot leveled off. Therefore, AcINH was analyzed by the acid hydrolysis method. Figures 7 and 8 show the standard calibration plots for INA and ING. The peak height ratio (PHR) is the ratio of the chromatogram peak height of the metabolite to that of an internal standard, PcINH. The points represent mean ± s.e. Table 2 displays summaries of the urinary recoveries of INA and ING (as % of total INH dose). The recovery of each metabolite (as % of total INH dose) was calculated as follows: % dose = Moles metabolite ÷ MOleS INH ingested X 100%. Moles IIVH ingested =0.3 g ÷ 137.15 (F.W. INH). Moles metabolite =Peak Height Ratio ÷ Standard Slope X Total volume of 24-hr urine (I) X 40 (dilution factor) X 10-6 (to convert from p, moles to moles). For example, % of dose for INA of subject ME (day 6) is calculated as follows: 0.352 (Peak Height Ratio) ÷ 0.0418 (slope, Fig. 7) X 1.136 I (total urine volume, Table 1) X 40 X i064- (0.3 g ÷ 137.15 Daltons) X i00%= 17.5% (Table 2).  56  Table 1. Summary of acetylator phenotype (R =rapid, s=slow), sex, and sample numbers of the 10 subjects from days 0, 6, and 7. Blank urine samples, 003E and 148M, are not included in the table. The volumes shown are for 24-hr urine collections. Day 0 (APAP Subject  Sex Acetylation  Day 6 (INH  )  Sample Volume (I) No.  )  Sample Volume (I) No.  Day 7 (INH+ APAP) Sample Volume (I) No.  ME  F  s  024E  1.640  063E  1.136  077E  0.550  JW  M  s  023W  2.258  064W  2.746  081W  2.815  JL  F  R  074L  1.540  112L  2.060  135L  1.527  DW  M  R  091DW 0.715  138DW 0.847  158DW  1.715  AC  M  s  139AC  3.265  186AC  3.955  251AC  3.920  MM  F  R  160M  0.840  219M  1.675  246M  1.060  DC  M  s  235DC  1.580  277DC  2.595  295DC  1.785  TZ  F  R  241Z  1.275  289Z  0.905  313Z  1.118  MC  F  s  252MC  1.673  290MC  0.465  330MC  1.210  SC  M  R  2695C  0.845  3125C  0.748  3355C  0.795  57  Figure 6. Chromatograms of urine samples analyzed at 0.5 A.U.F.S. A) 148M (blank), B) 160M (day 0: APAP), C) standard solution, D) 219M (day 6: INH), and E) 246M (day 7: APAP + INH). The compounds and their corresponding retention times (tR) are INA (peak 1, tR =3.89 min), AcINH (peak 2, tR discussed in text), ING (peak 3, tR =10.8 min), PcINH (peak 4, tR= 12.1 min) and interfering compound (peak 5, tR =9.84 min). Please see text for details.  58  Figure 7. Standard calibration plot for isonicotinic acid (INA). Each points  represent the mean (n =3) ± s.e.  59  Figure 8. Standard calibration plot for isonicotinylglycine (ING). Each points  represent the mean (n =3) ± s.e.  60 Table 2. Urinary recoveries (% of total dose of INH) of INA and ING on days 6 (INH) and 7 (INH + APAP). Values are the means (± S.E.M.) of a number of measurements (n). Subject  Sex  Acetylation  Day  n  INA  ING  Mean + S.E.  Mean + S.E.  ME  F  s  6 7  4 3  18 ±0.33 21 + 0.21  9.0 ±0.83 9.6 + 0.21  JW  M  s  6 7  3 3  18 + 0.18 18.+ 0.18  9.5 + 0.28 11 + 0.11  JL  F  R  6 7  3 4  24 + 0.13 22 + 0.48  16 + 1.1 14 +0.56  DW  M  R  6 7  3 4  19 + 0.35 24 + 0.42  11 + 0.20 14 + 1.3  AC  M  s  6 7  4 3  19 + 0.52 18 +0.57  15 + 0.91 16 +0.64  MM  F  R  6 7  3 3  25 + 0.36 32 ±0.37  16 ±0.83 18 ±0.23  DC  M  s  6 7  3 3  20 + 0.79 16 + 0.40  13 + 0.77 12 +0.42  TZ  F  R  6 7  4 3  24 ±0.52 23 +0.18  8.5 + 0.29 7.3 +0.43  MC  F  s  6 7  3 3  12 + 0.17 18 ±0.42  7.6 + 0.23 12 + 0.55  SC  M  R  6 7  4 4  15 + 0.32 23 + 0.25  7.8 + 0.23 12 + 0.35  61  3.2 RESULTS OF INH, ACHZ, AND HZ Fig. 9 shows sample chromatograms of the derivatized hydrazones of AcHz and INH, and the derivatized azine of Hz. The retention times (tR) and absorbance units full-scale (AUFS) are also shown. Figures 10, 11, and 12 show the standard calibration plots of the hydrazones of INH and AcHz, and the azine of Hz, respectively. The peak height ratio is the ratio of the chromatogram peak height of the derivatized metabolite to that of an internal standard, 9-fluorenone (9-Fl). The points represent mean ± s.e. Table 3 displays summaries of the urinary recoveries of INH, AcHz, and Hz (as % of total INH dose). The recovery of INH and each metabolite (as % of total dose of INH) was calculated as above. The dilution factors ranged from 1.33 to 10.  62  Figure 9. Chromatograms of standard solutions and retention times (tR) of the derivatized hydrazones of AcHz (1, tR =6.92 min) and INH (2, tR =7.76 min), and the azine of Hz (4, tR=40.4 min). 9-fluorenone (3, tR =16.5 min) is a volumetric internal standard. The large peak (located between peaks 2 and 3) belongs to m-anisaldehyde, the derivatizing compound. A sample chromatogram (0.32 A.U.F.S.) is shown on the left (sample 077E). The chromatogram (0.08 A.U.F.S.) of a solution containing 20 M N-acetyl-3-methoxybenzaldhydrazone (a measure of AcHz), 20 j M isonicotiny1-3-methoxybenzaldhydrazone (a measure of IN H), and 10 u M 3-methoxybenzaldiazine (a measure of Hz) is displayed on the right. The chromatogram of AcINH and DiAcHz is the same as that of the hydrazones of INH and AcHz, respectively.  63  y .0315x r=0.9998  0^50^100^150  ^  200  250  Conc. INH, pM  Figure 10. Standard calibration plot for isonicotiny1-3-methoxybenzaldhydrazone (a measure of INH). Each point represents the mean (n=3) ± s.e.  64  Figure 11. Standard calibration plot for N-acetyl-3-methmrybenzaldhydrazone (a measure of AcHz). Each point represents the mean (n=3) ± s.e.  65  0.30  0.20  0.10  y=.0337x + 0.0211  r=0.9990 0.00  a  2  4  6  8  Conc. Hz, pM  Figure 12. Standard calibration plot for 3-methoxybenzaldiazine (a measure of  Hz). Each point represents the mean (n=3) ± s.e. The intercept is most likely due to contamination; a standard sample contained AcHz, INH, and Hz. The AcHz was contaminated with a small amount of Hz.  66  Table 3. Urinary recoveries (% of total INH dose) of INH, AcHz, and Hz on days 6 (INH) and 7 (INH + APAP). Values are the means (± S.E.M.) of a number of measurements (n). Subject Sex Acetylation Day  n  INH  AcHz  Hz  Mean + S.E.  Mean + S.E.  Mean + S.E.  ME  F  s  6 7  3 3  20 + 0.30 31 + 0.42  3.7 + 0.053 5.0 + 0.044  0.70 + 0.030 0.74 + 0.039  JW  M  s  6 7  3 3  26 + 0.15 28 + 0.53  3.8 + 0.10 4.1 + 0.082  0.70 + 0.012 1.1 + 0.027  JL  F  R  6 7  3 3  7.8 + 0.076 6.5 +0.080  2.4 + 0.046 2.0 + 0.067  0.22 + 0.082 0.26 + 0.014  DW  M  R  6 7  3 3  6.2 + 0.098 9.8 + 0.15  1.5 + 0.064 1.8 + 0.065  0.30 + 0.0060 0.64 + 0.0037  AC  M  s  6 7  3 3  29 + 0.49 26 + 0.68  4.7 + 0.082 4.4 + 0.17  1.5 + 0.034 0.52 + 0.0085  MM  F  R  6 7  3 3  8.6 + 0.23 11 +0.23  2.4 + 0.071 2.6 + 0.020  0.56 + 0.012 0.41 + 0.012  DC  M  s  6 7  3 3  23 + 0.57 20 + 0.73  4.1 + 0.086 2.8 + 0.12  0.50 + 0.018 0.34 + 0.017  TZ  F  R  6 7  3 4  14 + 0.86 13 + 0.19  2.7 + 0.21 3.1 + 0.094  0.31 + 0.015 0.099 + 0.0064  MC  F  s  6 7  3 3  28 + 0.21 37 + 2.0  3.2 + 0.061 5.7 + 0.35  0.38 + 0.0065 0.85 + 0.046  SC  M  R  6 7  3 3  7.4 -I- 0.042 12 + 0.37  1.6 + 0.020 2.4 + 0.092  0.11 + 0.0040 0.42 + 0.0081  67  3.3 RESULTS OF ACETYLISONIAZID AND DIACETYLHYDRAZINE The chromatogram of AcINH and DiAcHz is the same as that of INH and AcHz, respectively (Fig. 9). As mentioned already, the original INH and AcHz in the urine samples were eliminated. Then AcINH and DiAcHz were converted to 'INH' and 'AcHz', respectively. 'INK and 'AcHz' were then derivatized with manisaldehyde and analyzed (the same procedure as that of the orininal INH and AcHz). Because of large inter-day variation in peak heights (due to variations of a few seconds in the length of partial acid hydrolysis) standard plots were set up each day as the samples of a subject were run. Fig. 13 shows the standard calibration plots for subject JL. The peak height ratio is the ratio of the chromatogram peak height of the metabolite to that of an internal volumetric standard, 9-fluorenone (9-Fl). The points represent mean ± s.e. The slopes of the calibration plots for AcINH ranged from 7.78X10-3 to 1.00X10-2 (r= 0.9935 to 0.9995). The slopes of the calibration plots for DiAcHz ranged from 2.43X10-3 to 2.98X10-3 (r= 0.9531 to 0.9992). The dilution factors ranged from 1 to 5. The final results are shown in Table 4. The recovery of each metabolite (as % of total dose of INH) was calculated as above. Extraction with dichloromethane eliminated essentially all of the original INH. After extraction, the residual INH peak was not larger than that of the baseline noise. The hydrolysis of DiAcHz generated AcHz. The concurrent hydrolysis of AcINH also generated a small amount of AcHz as a by-product. However, the peak height of this by-product was not larger than that of the baseline noise.  68  • AcINH  A DiAcHz  3.00  2.50  2.00  y= .00895x  r 0.9995 -  1.50  1.00  0.50  y=.00281x  r=0.9922 0.00 0^50^100 150 200 250 300 350 Concentration of AcINH or DiAcHz, pM  Figure 13.^Standard calibration plots for acetylisoniazid (AcINH) and diacetylhydrazine (DiAcHz). Each points represent the mean (n=3) ± s.e.  69 Table 4. Urinary recoveries (% of total dose of INH) of AcINH and DiAcHz on days 6 (INH) and 7 (INH + APAP). Values are the means (± S.E.M.) of a number of measurements (n). Subject  Sex  Acetylation  Day  n  AcINH  McHz  Mean + S.E.  Mean + S.E.  ME  F  s  6 7  4 3  11 + 0.18 14.3 + 0.719  3.6 + 0.052 5.5 + 0.25  JW  M  s  6 7  3 3  24 + 0.32 24 + 0.26  4.8 + 0.033 4.9 + 0.059  JL  F  R  6 7  3 4  63 + 1.2 49 + 0.43  38 + 0.64 25 + 0.092  DW  M  R  6 7  3 4  50 + 1.1 57 + 1.2  22 + 0.44 27 + 0.12  AC  M  s  6 7  4 3  28 + 0.70 26 + 0.47  8.1 + 0.11 6.5 + 0.094  MM  F  R  6 7  3 3  32 + 0.83 38 + 1.4  26 + 0.29 34 + 1.1  DC  M  s  6 7  3 3  38 + 0.53 37 + 0.27  9.3 + 0.22 9.5 + 0.033  TZ  F  R  6 7  4 3  68 + 0.36 60 + 0.84  25 + 0.020 20 + 0.13  MC  F  s  6 7  3 3  24 + 0.38 16 + 0.25  3.0 + 0.12 2.8 + 0.11  SC  M  R  6 7  4 4  38 + 0.91 62 + 0.72  13.0 + 0.31 29 + 0.55  70  3.4 SUMMARY OF RESULTS AND STATISTICS 3.4.1 RESULTS OF PAIRED COMPARISON STATISTICS  Table 5 shows a summary of recovery of the sum of isonicotiny (INH, AcINH, INA, and ING) and hydrazide (INH, AcINH, Hz, AcHz, and DiAcHz) compounds. Some recoveries were apparently greater than 100%. This is probably due to the fact that the standard errors are additive as the % recovery of various compounds were pooled together. The recoveries of the 7 compounds in rapid and slow acetylators are displayed in Table 6. The recoveries of all the metabolites for slow and rapid acetylator on day 6 (INH only) are similar to those reported in the literature. This study confirmed previous findings that the recovery of Hz and AcHz, suspected hepatotoxins, are greater in slow acetylators than rapid acetylators. The recovery of all the hydrazide compounds is greater in rapid acetylators (86% recovered) than slow acetylators (61% recovered). This suggests that more of INH undergoes microsomal metabolism in slow acetylators as compared to fast acetylators. The parametric two-tailed paired-sample t test was used to analyze the data. The differences were calculated within each pair of measurements (day 6 day 7). The paired-sample t test does not have the normality and equality of variances assumptions, but assumes only that the differences, di, come from a normally distributed population of differences. A histogram or stem-and-leaf display of the d's can provide a rough check for approximate normality. Our sample number (n =10) is too small for a stem-and-leaf plot. Histograms were prepared, but it was difficult to tell if the differences were normally distributed. Scatter plots were prepared that displayed distribution of differences. Figure 14 displays a log plot for the differences of the sum of isonicotinyl and hydrazide compounds. In order to do a log plot, the differences were all transformed linearly  71  by adding 100. Such coding does not change the measures of dispersion, except for the coefficient of variation. Figure 14 shows that it is difficult to tell if the differences are normally distributed or not. Alternatively, a nonparametric analogue to the paired-sample t test (the Wilcoxon paired-sample rank test) could be used. This test does not have the assumption of normality of differences, but it has an underlying assumption that the sampled population is symmetrical about the median (Zar, 1984). Two boxplots (Figs. 15 and 16; discussed later) show that the data are not symmetrical about the median. Another nonparametric test for paired samples is the sign test, one of ther oldest statistical procedures. However, the rank test is less powerful than the Wilcoxon test (Zar, 1984). The Wilcoxon test could not be applied because subgroup analysis (slow or rapid acetylators, n = 5) could not be done due to the small sample number. Since, it could not be assumed that the differences come from a normal distribution, we decided to do a log transformation of the original data before doing the parametric paired-sample t test. It was decided that if anything was significant by the parametric test, then it would be backed-up by the Wilcoxon test for n =10 (all subjects, slow and rapid acetylators). The null hypothesis was that concomitant ingestion of APAP would not alter the recoveries of INH and its metabolites. P < 0.05 was accepted as an indication of significance. There was no need to do Bonferroni corrections since nothing was statistically significant. Also, the 24-hr urinary creatinine recoveries did not change significantly from day 6 (1443 ± 140 mg) to day 7 (1510 ± 110 mg). Although treatment with APAP was associated with increased recoveries of most metabolites in fast acetylators, the effect was not statistically significant (Table 7). In slow acetylators, treatment with APAP was associated with  72  increased recoveries of some metabolites, but not all. However, the effect was not statistically significant. We are allowed to pool the results of the 5 slow and 5 fast acetylators (n=10) and then do a paired comparison. Although the paired comparison t-test  is the traditional method of solving this type of problem, Sokal & Rohlf (1987) recommend the two-way anova (paired comparison, randomized complete blocks design with only two treatments (a = 2)). The second test has the advantage of providing a measure of the variance component among blocks (b = 10). This is useful knowledge because if there is no significant added  variance component among blocks one might simplify the analysis and design of future, similar studies by employing a completely randomized anova. Failure to allow for differences among individuals can lead to erroneous results about the differences in INH metabolism on days 6 and 7. The two-way anova indicates that the individuals (b = 10) are different from each other (i.e., as expected, slow and fast acetylators are different from each other), but APAP has no effect on the metabolism of INH. However, concomitant ingestion of APAP was apparently not associated with changes in INH metabolism. Table 7 reports the calculated t-values (t s ) from the paired comparison t-test. It is known that t s 2 = Fs (Fs = calculated F-value of two-way anova paired comparison, i.e., randomized complete blocks design). The recoveries of INA (19% and 6% increase in fast and slow acetylators, respectively) and ING (8% and 9% increase in fast and slow acetylators, respectively) were increased upon concomitant ingestion of APAP, but these effects were not statistically significant. INA and ING both belong to the same metabolic pathway. When the results of INA and ING are pooled together (n=10), the t-test for paired comparisons indicates that APAP administration was not associated with a significant change in the conversion of AcINH and INH to INA/ING.  73 Table 5. Summary of recovery of total isonicotinyl (sum of INH, AcINH, INA, and ING) and total hydrazide (sum of INH, AcINH, Hz, AcHz, and DiAcHz) compounds. The values over 100% are most likely due to the additive effects of coefficients of variations. Subject  Sex  Acetylation  Day  Total Isonicotinyls  Total Hydrazides  (% of dose)  (% of dose)  ME  F  s  6 7  58 76  39 57  JW  M  s  6 7  78 80  59 62  JL  F  R  6 7  110 92  111 82  DW  M  R  6 7  86 104  80 96  AC  M  s  6 7  91 85  72 63  MM  F  R  6 7  81 100  70 86  DC  M  s  6 7  94 84  75 69  TZ  F  R  6 7  115 104  110 96  MC  F  s  6 7  71 83  58 62  SC  M  R  6 7  69 108  60 105  74 Table 6. Effect of APAP on the hepatic conversion of INH to its metabolites in 5 rapid acetylators and 5 slow acetylators. Values (mean + S.E.) show urinary recoveries (% of dose) on days 6 (IN H) and 7 (INH + APAP). Compound  Rapid Acetylators  Slow Acetylators  Day 6  Day 7  Day 6  Day 7  Isoniazid (INH)  8.8 + 1.4  10 + 1.1  25 + 1.7  28 + 2.7  Acetylisoniazid (AcINH)  50 + 6.9  53 + 4.4  25 + 4.4  23 + 4.0  Isonicotinic acid (INA)  21 + 1.8  25 + 1.8  17 + 1.4  18 + 0.84  Isonicotinylglycine (ING)  12 + 1.8  13 + 1.8  11 + 1.3  12 + 1.0  Diacetylhydrazine (DiAcHz)  25 + 4.0  27 + 2.3  5.7 + 1.3  5.8 + 1.1  2.1 + 0.25  2.4 + 0.24  3.9 + 0.25  4.4 + 0.48  0.30 + 0.075  0.36 + 0.090  0.75 + 0.19  0.72 + 0.14  Total isonicotinyls1  92 ±8.8  102 + 2.7  78 + 6.6  82 + 1.7  Total hydrazides2  86 + 10  93 + 4.0  61 + 6.3  63 + 2.0  Acetylhydrazine (AcHz) Hydrazine (Hz)  1 Sum of INH, AcINH, INA, and ING. 2Sum of INH, AcINH, Hz, AcHz, and DiAcHz.  75  A A  +  +  100 0  •  ^  0  •  30  Fast (I) Slow (I) Fast (H) Slow (H) Phenotype (compound)  Figure 14. Log plot of the distribution of differences (between days 6 and 7) of the sum of isonicotinyl (I) or hydrazide (H) compounds in slow or fast acetylators. The raw data were coded by adding 100. The 10 different symbols represent the 10 subjects. Each subject is represented by the same symbol in Figure 14, and Figures 17-20.  76 Table 7. Statistical results. The parametric paired-sample two-tailed t test was used to examine the effect of APAP on the conversion of INH to its metabolites in slow acetylators (n=5), rapid acetylators (n=5), or all the subjects (5 rapid and 5 slow acetylators, n=10). The values show the calculated t-values. The critical t-values are 2.78 (n=5), and 2.26 (n =10). Bonferroni corrections were not necessary since none of the calculated t-values were significant. Acetylation  Compound  All 10 Subjects Rapid  Slow  Isoniazid (INH)  -1.37  -0.96  -1.74  Acetylisoniazid (Acl NH)  -0.627  0.470  -0.184  Isonicotinic acid (INA)  -1.73  -0.448  -1.47  Isonicotinylglycine (ING)  -0.947  -1.22  -1.22  INA/ING  -1.48  -0.770  -1.64  Diacetylhydrazine (DiAcHz)  -0.604  -0.302  -0.709  Acetylhydrazine (AcHz)  -1.41  -0.590  -1.32  Hydrazine (Hz)  -0.368  0.0494  -0.277  Total Isonicotinyls1  -0.995  -0.780  -1.32  Total Hydrazides2  -0.713  -0.621  -0.979  1Sum of INH, AcINH, INA, and ING. 2Sum of INH, AcINH, Hz, AcHz, and DiAcHz.  77  3.4.2 GRAPHICAL ANALYSIS AND OTHER METHODS  Since the results of the paired-sample tests were not significant, attempts were made to obtain some information from i) boxplots, ii) scatterplots, iii) Hotelling T 2 statistics, and iv) calculations of the power of the tests and sample sizes required.  3.4.2.1 Box Plots At this stage of analysis it would be useful to have summary displays of the distribution. One method of summarization is by preparing box plots. A box plot allows a partial assessment of symmetry. If the distribution is symmetric then the box plot is symmetric about the median: the median cuts the box in half and the upper and lower vertical lines are about the same length. The upper and lower quartiles of the data are portrayed by the top and bottom, respectively, of the box. The median is displayed by a horizontal line segment within the box; it shows the centre of the distribution. The spread of the bulk of the data (the central 50%) is seen as the length of the box. The lengths of the vertical lines relative to the box show how stretched the tails of the distribution are (Chambers et al., 1983; Wilkinson, 1990a). Fig. 15 displays box plots of total recoveries of isonicotinyl compounds on days 6 (INH) and 7 (INH + APAP) in 5 fast and 5 slow acetylators. The data are not symmetrical about the median. The upper components of the box plot for day 6 data (fast and slow acetylators) are stretched relative to their counterparts below the median, revealing that the distribution is skewed to the right. Conversely, the distributions from day 7 are skewed to the left. None of the tails is symmetric. It is also apparent that the distribution of data of fast and slow acetylators are quite different from each other: this is due partly to the fact that recoveries of isonicotinyl compounds is greater in fast than slow acetylators  78  (Table 6). The recoveries of isonicotinyl compounds increase on day 7 compared to day 6; the increase is more drastic in fast acetylators. The variabilities of the day 7 distributions are drastically small compared to the respective day 6 distributions. Thus concomitant APAP ingestion is associated with reduction in the variation of the distributions. It should be noted that the sample size (n = 5) was small and that we would quite likely see different behavior in larger sample sizes. Fig. 16 displays box plots of total recoveries of isonicotinyl compounds on days 6 (IN H) and 7 (INH + APAP) in 5 men and 5 women on days 6 and 7. Again, the distributions are not symmetric. The variation of the distribution of the data of women is reduced on day 7 compared to day 6. Interestingly, the corresponding variation of the data of men is not reduced on day 7. These could be due to chance effects. These could also reflect the fact that the proportions of fast/slow acetylators in men and women were not equal. It is obvious the sex is not a major determinant of INH metabolism but acetylation phenotype is (this was also shown by variance statistics, discussed in section 3.4.2.2). The box plots of total recoveries of hydrazide compounds (not shown) are very similar to those of total isonicotinyl compounds. The box plots of individual metabolites were similar to those of the sums.  3.4.2.2 Scatter Plots and Variance Statistics Fig. 17 shows scatterplots (log) of total recoveries of isonicotinyl compounds in slow and fast acetylators on days 6 and 7. It is difficult to tell if the data are normally distributed or not. However, it is obvious that the variations of the distributions decrease on day 7 compared to day 6 (as shown also by the box plots). Fig. 18 confirms Fig. 16: on day 7 the variations of the distributions decrease in women but not in men.  79  Fig. 19 shows scatterplots (log) of total recoveries of hydrazide compounds in slow and fast acetylators on days 6 and 7. Again the variations decrease on day 7 compared to day 6. Fig. 20 (like Fig. 18) shows that the variability of distributions of female data decreases on day 7 but not those of males. Two-sample hypotheses were used to test the differences between variances (Zar, 1984). The results of day 6 or 7 were compared: 5 male versus 5 female subjects, and 5 slow versus 5 fast acetylators. The test assumes that the data are normally distributed. As discussed above, we are not able to show that the data are normally distributed. As mentioned above, the variation of the distribution of the data of women looks different on day 7 compared to the corresponding variation of the data of men. Subgroup analysis showed male versus female comparisons were not statistically significant (2-tailed test). But based on the small sample size, it is difficult to reach a definite conclusion. Previous studies have shown that the recoveries of INH, AcINH, Hz, AcHz, and DiAcHz are different between slow and fast acetylators but not between men and women. It was also shown before that the recoveries of INA and ING are similar in fast and slow acetylators, and men and women. Our study confirms these previous findings (Table 6). We could test for difference between two means but this is only done when the data is normally distributed. Therefore, it makes no sense to test these since our data do not look normally distributed.  3.4.2.3 Hotelling Multivariate Statistics When we test simultaneously the hypothesis that several population means do not differ from a specified set of constants, a statistic that considers all variables together is required. The multivariate Hotelling's T2 statistic (one-sample test) is usually used for this purpose (Wilkinson, 1990b; Zar, 1984). The test assumes  80  that the several variables are dependent on each other (the recovery of INH and its metabolites are dependent on each other). The differences of the original data were taken for days 6 and 7. The hypothesis was tested that the differences for all isonicotinyl compounds (or all hydrazide compounds) do not differ from zero (i.e., were equal). The test was also performed for INA and ING since they both  belong to the same metabolic pathway. In all cases there was no statistical significance.  3.4.2.4 Power of Test and Sample Size The 95% confidence intervals (original data) of the differences (between days 6 and 7) were computed for INH and each of its metabolites in slow acetylators, fast acetylators, or slow and fast acetylators combined (n=10). In all cases zero was within the computed interval. Therefore, it is not surprising that nothing was statistically significant by the paired-sample t test. By considering the paired-sample t test to be a one-sample t test for a sample of differences, cli, we can address the questions of minimum detectable difference, power, and required sample size (Zar, 1984). Table 8 displays the minimum detectable differences based on a sample size of 10 (5 Slow + 5 fast acetylators), or 5 (5 slow or 5 fast acetylators). An estimate was made of the smallest difference (i.e., difference between recoveries of each metabolite on days 6 and 7) that is detectable 80% of the time using a sample size of 10 or 5 data and a significance level of 0.05. With a sample size of 10, almost all the minimum detectable values are unacceptable. The test could detect about a 20% change in the recovery of INH and sum of isonicotinyl compounds with a reasonable degree of accuracy. In the case of Hz, an 80% change in recovery must occur so that the test can detect the change. When one looks at slow or fast acetylators separately, one expects homogeneity to increase  81  and variation to decrease. Therefore we expect the minimum detectable differences to decrease. However the estimations do not generally support this (probably due to the smaller sample sizes of 5). Table 9 shows calculations of the power of the test. A power of 0.80 is considered reasonably good. In agreement with Table 8, the power was 0.80 or more only in the case of ING and sum of isonicotinyl compounds. In the case of Hz, AcHz, and DiAcHz, there was no power at all. The required sample sizes were also estimated (Table 10). Only ING and sum of the isonicotinyl compounds required a sample size close to 10 (our situation). If we had equal numbers of slow and fast acetylators, we would need about 25 subjects to reject the null hypothesis. However proper detection of the recoveries of DiAcHz and Hz would require about 52 and 130 subjects, respectively. Alternatively, we would expect to reduce variability by studying only slow or fast acetylators. The required sample size of only slow or fast acetylators should be less than the situation where we have a mixture. However the estimates (Table 10) are not in full agreement; this is probably due to the fact that the estimates were based on data of only 5 subjects. It appears that if we had 25 slow acetylators, we could make definite conclusions about the study. The exception is the case of Hz: about 116 subjects may be required to draw a definite conclusion.  82  120  1  110  -  100  -  co 0 0 90 -0 i z ,_ 80 0 cx). 70  60  T  I^1^I  -  T -  T -  -  -  ^  -  -  ^  -  ■•••■■  50  1  Fast (6) Fast (7) Slow (6) Slow (7) CLASS  Figure 15. Box plots of total recoveries of isonicotinyl compounds in fast and  slow acetylators on days 6 (INH) and 7 (INH + APAP). The top and the bottom borders of each box show the 75% quartile and the 25% quartile, respectively, of the distribution. The median is shown by the horizontal line segment within each box. The data are not symmetrical about the median. In addition, the tails are not symmetric since the two vertical lines of each box are not of equal length. The variations of distributions decrease on day 7 compared to day 6.  83  120  1^ I ^ I  110  -  100 ED Cl) o -o T  -  T  T  -  T  90  -  -  7  ,_ 80 o cx) 70  60  50  I  -  -  ^  .■•■  -  -  1 ^1 ^ I  F (6)^F (7)  ^  V (6)  ^  V (7)  SEX  Figure 16. Box plots of total recoveries of isonicotinyl compounds in women (F) and men (M) on days 6 (INH) and 7 (INH + APAP). The top and the bottom  borders of each box show the 75% quartile and the 25% quartile, respectively, of the distribution. The median is shown by the horizontal line segment within each box. The distributions are not symmetric. On day 7 the variation of the distribution of the data of women is reduced, but not those of men.  84  +  CO  o  _  A  V 0  i +  e  • +  40 Fast (6) Fast (7) Slow (6) Slow (7) Phenotype (day)  Figure 17. Scatter plots (log) of total recoveries of isonicotinyl compounds in fast  and slow acetylators on days 6 and 7. The variations of distributions decrease on day 7 compared to day 6. The 10 different symbols represent the 10 subjects. Each subject is represented by the same symbol in Figure 14, and Figures 17-20.  85  a) o  •■•■••  O co 0 -0 = z ..... "6  1 00 -  .  + A^ • +^v o A^0  V  • 0^ • +  tl •  e  +  40  F (6)^F (7)^M (6)^M(7) Sex (day)  Figure 18. Scatter plots (log) of total recoveries of isonicotinyl compounds in females (F) and males (M) on days 6 and 7. The variations of distributions appear to decrease in women but not men. The 10 different symbols represent the 10 subjects. Each subject is represented by the same symbol in Figure 14, and Figures 17-20.  86  _  g,^100  -.7_--0 co 0 -0 = Z — 0 e  _  4  • + 0 A  a^ o 0  i  It^+  + 30 Fast (6) Fast (7) Slow (6) Slow (7) Phenotype (day)  Figure 19. Scatter plots (log) of total recoveries of hydrazide compounds in fast and slow acetylators on days 6 and 7. The variations of distributions decrease on day 7 compared to day 6. The 10 different symbols represent the 10 subjects. Each subject is represented by the same symbol in Figure 14, and Figures 17-20.  87  4  •  +  -  0 A  o  V V 0 o  0  + 30  F (6)^F (7)^M (6)^M (7) Sex (day)  Figure 20. Scatter plots (log) of total recoveries of hydrazide compounds in females (F) and males (M) on days 6 and 7. The variations of distributions appear  to decrease in women but not men. The 10 different symbols represent the 10 subjects. Each subject is represented by the same symbol in Figure 14, and Figures 17-20.  88 Table 8. Minimum percent change detectable. The minimum value of 8 (the difference between the mean values of recoveries on days 6 and 7 for each metabolite), that is detectable 80% of the time by the paired-sample t test at the a (0.05) level of significance, using a sample of specified size n. A 8 was calculated by the formula below (Zar, 1984). A 5 was then divided by the mean value of each metabolite (day 6) to get the minimum percent change detectable. The mean % change observed is the % of increase (or decrease) of the recovery of each metabolite on day 7 compared to day 6.  Compound  ^ ^ All Subjects Rapid Acetylators Slow Acetylators (n=10) (n=5) (n=5) Min. % Mean^Min. % Mean Mean^Min. % % Change^Change % Change^Change % Change Change Observed^able to Observed able to Observed^able to Detect Detect Detect  INH AcINH  14 1.6  27 27  INA ING  11 9.7 7.2  22 20  13 2.9 7.4  5.9  DiAcHz AcHz Hz Total isonicotinyls (I INH, AcINH, INA,  51 48 34 36 76  12 -6.8 4.0  34 80 20  19 5.7 16 9.8 8.9 14 21 10  36 42 42  12 -4.6 4.2  62 128 25  28  8.1  55  3.1  28  50  10 1.1  42 27 39 33 37  and ING)  Total hydrazides (E INH, AcINH, Hz, AcHz, and DiAcHz) s  5 =--v71-(tay + towy ) s =standard deviation of the differences (day 6 - day 7) of the original data tocy:t 0.05(2),9= 2.26, t 0.05(44=2.78 t 0 toy : t o.20(1),9 =0.880, t 0.2o(11,4= 0.941  89 Table 9. Power of the paired-sample t test. The probability of detecting a true difference, i.e., a difference between mean recoveries on day 6 and those on day 7 of at least 20%. The value of 8 (the difference between the mean values of recoveries on days 6 and 7 for each metabolite), is arbitrarily set to 20%; i.e., 8 is equal to 20% of the recoveries on day 6. a =0.05, n = 10, and t 0(11,9 was calculated as shown below. The corresponding p values were found from a statistical table, and the power (1- 0) was calculated. Mean  0 Values  Recovery Day 6 (% INH dose)  Calculated 8  Standard Deviation, s  Calculated t p (1),9  INH  17  3.4  4.7  0.050  0.48  0.52  AcINH  38  7.5  10  0.040  0.48  0.52  INA  19  3.9  4.2  0.65  0.26  0.74  ING  11  2.26  2.2  0.94  0.17  0.83  DiAcHz  15  3.1  7.7  -1.0  AcHz  3.0  0.60  1.02  -0.40  0.53  0.11  0.42  -1.48  85  17  17  0.85  0.20  0.80  74  15  20  0.01  0.50  0.50  Compound  Hz Total isonicotinyls  (Statistical Table)  Power of Test  (1-0)  None -  None None  (E INH, AcINH, INA, and ING)  Total hydrazides ( I INH, AcINH, Hz,  AcHz, and DiAcHz)  Jet  4  t (30),A, =8 x - - tax S  S=standard deviation of the differences (day 6 - day 7) of the original data  tax: t  0.05(2),9 =  2.26,  90 Table 10. Sample size required. We want to be able to detect a difference between days 6 and 7 as small as 5 (20%). We wish to test at the 0.05 level of significance with a 80% chance of detecting means (day 7) significantly different from day 6, means by as little as 20%. The required sample size was calculated by rearranging the equation (Table 9) for n. For all subjects, the estimated minimum sample size was based on n=10 set of values; for slow or rapid acetylators, the estimated minimum sample size was based on n =5 set of values. All Subjects  Compound  Rapid acetylators^Slow Acetylators  Calculated Minimum^Calculated Minimum^Calculated Minimum Sample Size Required Sample Size Required Sample Size Required  INH  17  21  15  AcINH  17  19  8  INA  12  11  13  ING  10  12  10  - 52  - 43  12  25  12  21  -130  -142  -116  11  15  7  18  24  8  DiAcHz AcHz Hz Total isonicotinyls (E INH, AcINH, INA, and ING)  Total hydrazides (E INH, AcINH, Hz, AcHz, and DiAcHz)  91  4 DISCUSSION We previously reported that INH inhibits the oxidative metabolism (minor pathway) of APAP (Epstein et al., 1991). A more thorough study suggests that the effect of INH on the metabolism of APAP is actually quite complex (Zand eta!., 1992). Depending on the condition, INH inhibits or induces P450 2E1, the major enzyme that catalyzes the toxic pathway of APAP metabolism.  Cases of APAP hepatotoxicity have been reported following the ingestion of long-term INH and high doses of APAP (Moulding eta!., 1991; Murphy eta!., 1990). The hepatotoxicity was most likely due to induction of P450 2E1 metabolism of APAP by isoniazid (Zand eta!., 1992).  At doses of 1.6 g/day or more, APAP may inhibit the microsomal oxidative pathway of warfarin, but this has not been substantiated (Boeijinga et al., 1982; Bartle & Blakely, 1991). APAP competitively inhibits the sulfation of ethinyl estradiol, an oral contraceptive (Rogers et al., 1987). APAP has been reported to increase, decrease, and have no effect on the clearance of chloramphenicol. Intravenous, but not oral, doses of APAP might competitively inhibit the glucuronidation of chloramphenicol (Spika & Aranda, 1987; Stein et al., 1989). The reasons for the disparate results in the chloramphenicol studies probably result from differences in study design, number of subjects, assay techniques, or route of APAP administration. In addition, the clearance of chloramphenicol has been reported to increase with continuous dosing. The clinical significance of this interaction is still speculative.  The present study suggests that a low dose of APAP (500 mg) has very little effect on INH metabolism. However, due to the small sample size and great  92  variability in the metabolism of INH we can not draw a definite conclusion. APAP administration was associated with increases in the conversion of INH to some of its metabolites, but the effects were not statistically significant. This will not result in an adverse drug interaction. For example, the recoveries of INA and ING increased, but these two metabolites are relatively nontoxic; the effect of APAP, if any, represents a clinically inconsequential drug interaction. If APAP does in fact have an effect on the metabolism of INH, we were not able to detect it. A high dose of APAP may alter the metabolism of INH significantly, but it is unethical to conduct such a study because of the higher risk of liver toxicity. Just like the case of APAP/warfarin, a low dose of APAP may have no effect, whereas a higher dose might have. There are some possible explanations for the results of this study.  Due to our sample size and great variability in the metabolism of INH in humans, the power of our paired-sample t-test was generally not satisfactory. In order to be able to do proper statistics, either the sample size would have to be increased, or variability decreased; e.g., all slow acetylator or all fast acetylators. Perhaps with a larger sample size (e.g., 21 slow acetylators) the data will be normally distributed. The recoveries of INH and its metabolites in slow acetylators are more homogeneous than those in fast acetylators. To increase homogeneity, fast acetylators could be classified into 2 groups with acetylator genotyping. However, as Table 8 (p. 88) shows, this study still provided some useful information. For example, we can be reasonably confident that APAP does not cause a change of 42% or more in the metabolism of INH. Similarly, the urinary recovery of AcINH would not be changed by more than 27% (Table 8). We can be fairly confident that APAP does not cause a change of 50% or more in the recoveries of INH and most of its metabolites in fast or slow acetylators (Table 8).  93  The mean recovery of isonicotinyl compounds is higher by 15-20% in fast acetylators than slow acetylators (Table 6, p. 74). Ideally, the recoveries should be identical in slow and fast acetylators. There are many possible explanations. First, it is possible that more of INH is converted to other isonicotinyl compounds (e.g., DiINH, 4-hydroxymethylpyridine, etc.) in slow acetylators than fast acetylators. This appears to be a reasonable explanation, considering the fact that the levels of INH are higher in the body of slow acetylators compared to fast acetylators. In this study, DiINH and 4-hydroxymethylpyridine were not analyzed. Literature reports suggest almost equal recoveries of isonicotinyl compounds in slow and fast acetylators (Ellard & Gammon, 1976).  Second, it is possible that the determinations of AcINH by the acid hydrolysis method were not very accurate. For example, five times out of six the method could accurately determine the concentration of AcINH present in spiked urine. But one time out of six the determination was off by 14%. It is quite possible that overestimation of AcINH levels in some urine samples and underestimation of others have led to different estimates in slow and fast acetylators. If so, then the recoveries of hydrazide compounds are very different due to the same reason. This artefact of methodology obviously would have major effects on the distributions of the box plots and scatter plots. The second reason is more likely; this is supported by the fact that the the estimated recoveries of isonicotinyl compounds were over 100% for some subjects (Table 5, p. 73).  One positive finding of this experiment is that APAP causes a larger reduction in the variability of recoveries (of isonicotinyl compounds) in fast acetylators as compared to slow acetylators (Figure 15, p. 82). If this finding is  94  not caused by any artefact of methodology, then it would suggest an interaction between APAP and the enzyme N-acetyltransferase (NAT). High levels of AcINH inhibit NAT. Since APAP has an acetyl group like AcINH, it would be expected to inhibit NAT. Furthermore, the inhibition should be most dramatic in the fastest fast acetylator subject (i.e., the faster the acetylation rate, the more inhibition). However, the results of AcINH recoveries in Table 4 (p. 69) are conflicting. The recovery AcINH in subject TZ, the fastest of fast acetylators, decreases by 13% when INH and APAP are taken together. The recovery of AcINH in subject JL, the second fastest of fast acetylators, decreases by 22%. The corresponding recoveries in subject DW, the third fastest of fast acetylators, increases by 14%. Therefore, there does not appear to be a trend.  Although calculation of intrinsic metabolic clearance is the preferred means of assessing drug metabolizing enzyme activity in vivo, simpler but indirect measures of metabolic capacity are often utilized (Miners et al., 1992). One example is the urinary metabolic ratio. For example, in the study of carbamazepine-induced INH hepatotoxicity, it was shown that carbamazepine decreased the ratios, DiAcHz/AcHz, AcINH/INH, and ING/INA, by 37%, 31% and 63%, respectively (Wright et al., 1984). In our present study, the corresponding ratios were decreased by 5.4%, 6.7%, and 8.8% in fast acetylators; none of the ratios changed that drastically. If APAP had drastic effects, e.g. 50% change in recoveries of any of INH, AcINH, INA, ING, AcHz, or Hz in fast acetylators, the paired-sample t-test would have been powerful enough to detect the changes (Table 8, p. 88).  In a follow-up clinical study, the effect of different doses of APAP (e.g., 325 mg, 650 mg, and maybe 1000 mg) could be tested on the conversion of INH to  95  AcINH in 21 fast acetylators. The effect of APAP on the pharmacokinetic parameters of INH could also be investigated. Alternatively, basic studies could be done with rabbits. In addition, in vitro enzyme studies could be conducted.  If APAP had shown significantly important effects on the metabolism of INH, there could be many possible explanations. First, the significance could have been due to a chance effect. Second, INH could have altered its own metabolism: i) there could have been accumulation of some of the metabolites in the body due to long half-lives and subsequent increased clearance by the kidneys. For example, Blair et al. (1985) administered INH (300 mg) daily for 14 days to 4 slow and 4 rapid acetylators. They reported that accumulation of Hz occurred in slow acetylators by the end of the study. The authors did not suggest any possible mechanisms. INH might have inhibited the P-450 metabolism of Hz, and subsequently caused the levels of Hz to rise in the blood. As the subjects continued to ingest more INH every day, accumulation of Hz occurred. If the levels of Hz continue to rise in the blood, then the clearance of Hz by the kidneys would be expected to increase. Beyond all this, it is known that Hz (and AcHz) have longer half-lives than INH. From our study we would not be able to tell if INH altered its own metabolism.  INH could have altered its own metabolism by another mechanism: ii) INH induced its own metabolism through P450 2E1 (Wall, personal communication). Microsonnal metabolism of both INH and AcINH could occur. Generally, a drug that induces a cytochrome P450 enzyme is also usually a substrate for it. Nhydroxylation of many compounds is carried out by P450 2E1. We believe (although not proven yet) that P450 2E1 hydroxylates the hydrazine nitrogen of INH adjacent to the isonicotinyl ring, and converts INH to INA and other by-  96  products (Wall, personal communication). In fact, it was just shown that myeloperoxidases (MPO)/H202 in activated neutrophils and monocytes convert INH to INA by oxidation, and it is likely that a radical intermediate is involved in this metabolic pathway (Hofstra eta!., 1992).  P450 2E1 is responsible for the metabolic activation of many chemicals and carcinogens (Hyland et al., 1992; Gonzalez, 1992). P450 2E1 is known to efficiently catalyzes the oxidation or reduction of more than 75 low-molecularweight substrates (Koop, 1992; Ingelman-Sundberg et al., 1992). INH was not listed as an oxidative substrate of P450 2E1, because no one has studied this.  Third, the renal clearance of some INH metabolites might have been increased by APAP administration. APAP, INH, and AcINH undergo glomerular filtration. INA, ING, APAP glucuronide, and APAP sulfate are all actively secreted by the kidneys. APAP (4.0 g/day for 3 days) had no significant effect on the glomerular filtration rate in healthy volunteers (Prescott et al., 1990). The renal clearance of INA and ING might have been increased by APAP administration, but we have no reason to believe so. APAP glucuronide and APAP sulfate might inhibit the clearance of INA and ING, not increase it.  Fourth, APAP might have interacted allosterically with the enzyme amidase and increased its activity. If this happened, we would expect increased recoveries of INA/ING, AcHz, and Hz. However, AcHz and especially Hz are very reactive, and they are rapidly converted to other products; we would not be able to tell if the activity of amidase increased. The only way to find out would be to do a direct enzyme study with amidase.  97  Fifth, APAP could potentially affect the metabolism of INH by interacting with amidase, NAT, or cytochrome P450(s). The enzyme amidase is probably involved in the hydrolysis of INH and AcINH. Studies in rats have shown that annidase catalyzes the hydrolysis of APAP to yield p-aminophenol, a minor pathway of APAP metabolism (Klos et al., 1992). Therefore APAP and INH are both substrates for amidase, but the latter is a much better substrate. A number of amidases are present in the mitochondrial, microsomal, and soluble fractions of cells. In vitro studies, using rat liver subcellular fractions, revealed that INH amidase activity was widespread in all liver subfractions; however, relatively high concentrations of the enzyme were localized in the liver microsomal and lysosomal fractions (Sendo et al., 1984). The INH amidase activity was inhibited extensively by BNPP, moderately by AcINH, and slightly by acetanilide. Because AcINH is also a substrate for amidase, the inhibition by AcINH was probably a competitive one. It is possible that the slight inhibition by acetanilide, a paminophenol derivative that gets converted to aniline by amidase, was a competitive one. Assuming that amidase also catalyzes the hydrolysis of APAP (another p-aminophenol derivative) in humans, there is a potential for drug interaction. But this would be unlikely because, at therapeutic doses, the hydrolysis of APAP is a minor pathway. It remains to be seen how low doses and high doses of APAP affect INH amidase in humans.  The cytosolic enzyme, polymorphic N-acetyltransferase (NAT; E.C. 2.3.1.5), catalyzes the acetylation of INH and its metabolites, Hz and AcHz. A microsomal NAT (cysteine S-conjugate NAT; E.C. 2.3.1.80) acetylates the APAP metabolite, 3-Cys-APAP, to APAP 3-mercapturate. This enzyme has been partially purified from rat kidney microsomes and appears to favor more hydrophobic thioethers (Duffel & Jakoby, 1982). Both liver and kidney contain the  98  enzyme with the specific activity in the kidney being about twice that in liver (Green & Elce, 1975). This microsomal NAT is distinct from the soluble NAT that catalyzes the N-acetylation of many endogenous and exogenus amines and is present in these and other tissues (Stevens & Jones, 1989). Microsomal NAT is inibited by the same agents that inhibit cytosolic NAT. Rat kidney microsomal NAT is unaffected by 1 mM GNI" , Mg2+, Ca2+, and Co2+ while Zn2+ and Mn2+ are moderately inhibitory and Cu2+ is totally inhibitory at that concentration (Weber et al., 1980). EDTA is slightly stimulatory but does not reverse Cu2 4. inhibition. N-ethylmaleimide and p-chloromercuribenzoate are both inhibitory at 1 mM concentration.  Studies on the metabolism of the glutathione S-conjugate of APAP by isolated rat kidney cells showed that these cells can take up the cysteine Sconjugate and N-acetylate it (Stevens & Jones, 1989). Liver and kidney probably represent most of the cysteine S-conjugate removal from the plasma. It is not known if uptake and N-acetylation of cysteine S-conjugates occurs in all other tissues. Apparently rat brain, spleen, stomach, heart and small intestine lack activity (Weber et al., 1980). This is an important concern given the recognition that cysteine S-conjugates can be toxic (Stevens & Jones, 1989).  Substrates effective with cytosolic NAT, e.g., aniline and p-aminobenzoic acid (PABA), were not acetylated by rat microsomal NAT (Duffel & Jakoby, 1982). In humans, PABA and aniline are preferentially acetylated by the monomorphic and polymorphic NAT, respectively. However, studies of genetic variation in Nacetylation in a rat model are few. In the rat liver, apparently a common NAT catalyzes the acetylation of PABA and INH. Other rat tissues such as lung, kidney, gut, and spleen also acetylate PABA (Weber, 1987). It appears that rat  99  NAT acetylates drugs monomorphically. However, there is a high variation in liver PABA NAT activity across inbred rat strains (Weber, 1987). Since there are no reports in the literature of studies on human microsomal NAT, it is not not known if INH is a substrate for this enzyme. Even if human microsomal NAT could acetylate INH, a drug interaction would not be expected since the acetylation of 3Cys-APAP is a minor pathway (at therapeutic doses of APAP).  APAP (4.0 g/d for 3 days) apparently inhibits the acetylation of pamminohippurate (PAH) in humans since it reduced the urinary excretion of the  metabolite while the recovery of the total PAH was unchanged (Prescott et al., 1990). Apparently the mouse kidney can rapidly acetylate PAH (Carpenter & Mudge, 1981); if the human kidney is also the primary organ where PAH is acetylated, then it would suggest the involvement of a monomorphic NAT. Conversely, acetanilide, a compound structurally related to APAP, is preferentially acetylated by polymorphic NAT in humans (Weber, 1987). Therefore, the interference of APAP with the acetylation of PAH suggests that the latter could be a substrate for polymorphic NAT, i.e., APAP interferes with polymorphic NAT. If this is the case, then high doses of APAP would be expected to affect the acetylation of INH.  Because APAP is metabolized by many types of P450 enzymes (P450 2E1, 3A4, 1A2, and perhaps other types in humans) it would have the potential to interfer with the P450(s) that metabolize INH, AcHz, and Hz. As mentioned already, APAP (doses of 1.6 g/day or more) may inhibit the microsomal oxidative pathway of warfarin; but this has not been substantiated. APAP, so far, has not been reported to affect the microsomal metabolism of any other drug.  100  Based on the results of this study, we can conclude that APAP does not have a major effect (e.g., 50% change) on the metabolism of INH. However, APAP might have small effects on the metabolism of INH. For example, APAP might interfere with the enzyme NAT. Based on two other clinical studies (Epstein et al., 1991; Zand et al., 1992) INH does influence the metabolism of APAP. However, no significant clinical drug interaction is expected between APAP and INH at therapeutic doses. It remains to be seen whether APAP has any effect on the metabolism of INH or not.  101  5 REFERENCES Albano, E., Rundgren, M., Harvison, P.J., Nelson, S.D. & Moldeus, P. (1985). Mechanisms of N-acetyl-p-benzoquinone imine cytotoxicity. Molec. PharmacoL, 28, 306-311. Albano, E., & Tomasi, A. (1985). Spin trapping of free radical intermediates produced during the metabolism of isoniazid and iproniazid in isolated hepatocytes. Biochem. PharmacoL, 36, 2913-2920. Alvares, A.P. & Pratt, W.B. (1990). Pathways of drug metabolism. In Principles of Drug Action, eds.: Pratt, W.B. & Taylor, P., 3rd ed., pp. 365-422. Andrews, R.S., Bond, C.C., Burnett, J., Saunders, A. & Watson, K. (1976). Isolation and identification of paracetamol metabolites. J. Mt. Med. Res., 4 (Suppl. 4), 34-39. Baciewicz, A.M. & Self, T.H. (1985). Isoniazid interactions. Southern Med. J., 78, 714-718. Bahri, A.K., Chiang, C.S. & TimbreII, J.A. (1981). Acetylhydrazine hepatotoxicity. Toxicol. App!. PharmacoL, 60, 561-569. Bahri, A.K., Chiang, C.S. & TimbreII, J.A. (1982). Factors affecting acetylhydrazine hepatotoxicity. Adv. Exp. Med. Biol., 136(B), 1055-1065. Barbare, J.C., Lallement, P.Y., Vorhauer, W. & Veyssier, P. (1986). Hepatotoxicite de l'isoniazide: influence de la carbamazepine? (lett.). GastroenteroL Clin. BioL, 10, 523-524. Bartle, W.R. & Blakely, J.A. (1991). Potentiation of warfarin anticoagulation by acetaminophen (lett.). JAMA, 265, 1260. Behnisch, R., Mietzsch, F. & Schmidt, H. (1950). ^Chemical studies on thiosemicarbazones with particular reference to antituberculous activity. Am. Rev. Tuberc., 61, 1-7. Bernstein, J., Lott, W.A., Steinberg, B.A. & Yale, H.L. (1952). Chemotherapy of experimental tuberculosis. V. Isonicotinic acid hydrazide (nydrazid) and related compounds. Am. Rev. Tuberc., 65, 357-364. Black, M., Mitchell, J.R., Zimmerman, H.J., Ishak, K.G. & Epler, G.R. (1975). Isoniazid-associated hepatitis in 114 patients. GastroenteroL, 69, 289-302. Black, M. (1984). Acetaminophen hepatotoxicity. Ann. Rev. Med., 35, 577-593. Blair, I.A., Mansilla-Tinoco, R., Brodie, M.J., Clare, R.A., Dollery, C.T., TimbreII, J.A. & Beever, I.A. (1985). Plasma hydrazine concentrations in man after isoniazid and hydralazine administration. Human ToxicoL, 4, 195-202. Boeijina, J.J., Boestra, E.E., Ris, P., Breimer, D.D. & Jeletich-Bastiaanse, A. (1982). Interaction between paracetamol and coumarin anticoagulants (lett.). Lancet, i, 506.  102  Boxenbaum, H.G. & Riegelman, S. (1976). Pharmacokinetics of isoniazid and some metabolites in man. J. Pharmacokin. Biopharmaceut, 4, 287-325. Boxenbaum, H.G., Jodhka, G.S., Ferguson, A.C., Riegelman, S. & MacGregor, T.R. (1974). The influence of bacterial gut hydrolysis on the fate of orally administered isonicotinuric acid in man. J. Pharmacokin. Biopharmaceut, 2, 211-237. Bozek, C.B., Kastrukoff, L.F., Wright, J.M., Perry, T.L. & Larsen, T.A. (1987). A controlled trial of isoniazid therapy for action tremor in multiple sclerosis. J. NeuroL, 234, 36-39. Brewer, G.A. (1977). Isoniazid. Analytical Profiles of Drug Substances, 6, 183-258. Burk, R.F., Hill, K.E., Hunt, R.W. & Martin, A.E. (1990). Isoniazid potentiation of acetaminophen hepatotoxicity in the rat and 4-methylpyrazole inhibition of it. Res. Commun. Chem. Path. Pharmacot, 69, 115-118. Carpenter, H.M. & Mudge, G.H. (1981). ^Uptake and acetylation of pamminihippurate by slices of mouse kidney cortex. J. PharmacoL Exp. Ther., 213, 350-354. Cash, J.M. & Zawada, E.T. (1991). Isoniazid overdose. Successful treatment with pyridoxine and hemodialysis. West. J. Med., 155, 644-646. CDC: Centers for Disease Control (1989). A strategic plan for the elimination of tuberculosis in the United States. MMWR, 38 (Suppl. no. S-3). Chambers, J.M., Cleveland, W.S., Kleiner, B. & Tukey, P.A. (1983). Graphical Methods for Data Analysis. Wadsworth International Group, Belmont, CA. Chorine, V. (1945). Action de l'annide nicotinique sur les bacilles du genre Mycobacterium. C.R. Acad. ScL (Paris), 220, 150-151. Clissold, S.P. (1986). Paracetamol and phenacetin. Drugs, 32 (Suppl. 4), 46-59. Colvin, L.B. (1969). Metabolic fate of hydrazine and hydrazides. J. Pharmaceut Sc., 58, 1433-1443. Critchley, J.A., Dyson, E.H., Scott, A.W., Jarvie, D.R. & Prescott, L.F. (1983). Is there a place for cimetidine or ethanol in the treatment of paracetamol poisoning? Lancet, i, 1375-1376. Cuthbertson, W.F.J., Ireland, D.M. & Wolff, W. (1953). Detection and identification of some metabolites of isonicotinic acid hydrazide (isoniazid) in human urine. Biochem. J., 55, 669-671. Dahlin, D.C., Miwa, G.T., Lu, A.Y. & Nelson, S.D. (1984). N-acetyl-p-benzoquinone imine: a cytochrome P-450-mediated oxidation product of acetaminophen. Proc. Natt Acad. ScL USA, 81, 1327-1331. Davis, M., Simmons, C.J., Harrison, N.G. & Williams, R. (1976). Paracetamol overdose in man: relationship between pattern of urinary metabolites and severity of liver damage. Quart. J. Med., 45, 181-191.  103 Dickinson, D.S., Bailey, W.C., Hirschowitz, B.I., Soong, S.J., Eidus, L. & Hodgkin, M.M. (1981). Risk factors for isoniazid (INH)-induced liver dysfunction. J. Clin. GastroenteroL, 3, 271-279. Dockweiller, D. (1987). Isoniazid-induced valproic-acid toxicity, or vice versa (lett.). Lancet, ii, 152. Domagk, G., Behnisch, R., Mietzsch, F. & Schmidt, H. (1946). Uber eine neue, gegen Tuberkelbazillen in vitro wirksame verbindungsklasse. Naturwissenschaften, 33, 315. Duffel, M.W. & Jakoby, W.B. (1982). Cysteine S-conjugate N-acetyltransferase from rat kidney microsomes. Mol. PharmacoL, 21, 444-448. El-Naggar, A.M., Hussein, M.E., El-Nemma, E.M. & Samour, A.A. (1985). Synthesis and antimicrobial activity of some new isonicotinylamino acid and dipeptide derivatives. H. Farmaco-Ed. Sc., 40, 662-670. Ellard, G.A. (1984). The potential clinical significance of the isoniazid acetylator phenotype in the treatment of pulmonary tuberculosis. Tubercle, 65, 211-227. Ellard, G.A. & Gammon, P.T. (1976). Pharmacokinetics of isoniazid metabolism in man. J. Pharmacokin. Biopharmaceut., 4, 83-113. Ellard, G.A., Gammon, P.T. & Wallace, S.M. (1972). The determination of isoniazid and its metabolites acetylisoniazid, monoacetylhydrazine, diacetylhydrazine, isonicotinic acid, and isonicotinyl-glycine in serum and urine. Biochem. J., 26, 449-458. Ellard, G.A., Mitchison, D.a., Girling, D.J., Nunn, A.J. & Fox, W. (1978). The hepatotoxicity of isoniazid among rapid and slow acetylators of the drug (lett.). Am. Rev. Resp. Dis., 118, 628-629. Epstein, M.M., Nelson, S.D., Slattery, J.T., Kalhorn, T.F., Wall, R.A. & Wright, J.M. (1991). Inhibition of the metabolism of paracetamol by isoniazid. Br. J. Clin. PharmacoL, 31, 139-142. Erill, S., Palacios, G., Costa, J. & Laporte, J.R. (1977). A comparison of the liver toxicity of isoniazid, cyanazide, and their acetyl-derivatives. In Drug Design and Adverse Reactions. Eds.: Bundgaard, H., Juul, P. & Koford, H. Alfred Benzon Symposium X, Copenhagen. Academic Press, New York., pp. 89-96. Fairbrother, J.E. (1974). Acetaminophen. Analytical Profiles of Drug Substances, 3, 1-109. Forrest, J.A.H., Clements, J.A. & Prescott, L.F. (1982). clinical pharmacokinetics of paracetamol. Clin. Pharmacokin., 7, 93-107. Fox, H.H. (1952). Synthetic tuberculostats: I. Pyridine carboxylic acid derivatives. II. Amino- and hydroxy-pyridine carboxylic acid derivatives. III. Isonicotinaldehyde, thiosemicarbazone and some related compounds. J. Org. Chem., 17, 542-562.  104  Fox, S.W. & Field, H. (1943). The synthesis of nicotinuric acid. J. Biol. Chem., 147, 651-652. Francis, D.A., Grundy, D. & Heron, J.R. (1986). The response to isoniazid of action tremor in multiple sclerosis and its assessment using polarised light goniometry. J. NeurobioL, 49, 87-89. Gardner, T.S., Wenis, E. & Lee, J. (1954). The synthesis of compounds for the chemotherapy of tuberculosis. IV. The amide function. J. Org. Chem., 19, 753-757. Gent, W.L., Seifart, H.I., Parkin, D.P., Donald, P.R. & Lamprecht, J.H. (1992). Factors in hydrazine formation from isoniazid by paediatric and adult tuberculosis patients. Eur. J. Clin. PharmacoL, 43, 131-136. Gershanik, 0.S., Luquin, M.R., Scipioni, 0. & Obeso, J.A. (1988). Isoniazid therapy in Parkinson's disease. Mov. Disord., 3, 133-139. Ghatineh, S., Morgan, W., Preece, N.E. & TimbreII, J.A. (1992). A biochemical and NMR spectroscopic study of hydrazine in the isolated rat hepatocyte. Arch. ToxicoL, 66, 660-668. Gonzalez, F.J. (1992). Human cytochromes P450: problems and prospects. T.LP.S., 13, 346-352. Grant, D.M., morike, K., Eichelbaum, M. & Meyer, U.A. (1990). Acetylation pharmacogenetics. The slow acetylator phenotype is caused by decreased or absent arylamine N-acetyltransferase in human liver. J. Clin. Invest, 85, 968972. Green, R.M. & Elce, J.S. (1975). Acetylation of S-substituted cysteines by a rat liver and kidney microsomal N-acetyltransferase. Biochem. J., 147, 283-289. Gregus, Z., Madhu, C. & Klassen, C.D. (1988). Species variation in toxication and detoxication of acetaminophen in vivo: A comparative study of biliary and urinary excretion of acetaminophen metabolites. J. Pharmacot Exp. Ther., 244, 91-99. Guengerich, F.P. (1989). Polymorphism of cytochrome P-450 in humans. T.LP.S., 10, 107-109. Gurumurthy, P., Krishnamurthy, M.S., Nazareth, 0., Parthasarathy, R., Sarma, R.G., Somasundraram, P.R., Tripathy, S.P. & Ellard, G.A. (1984). Lack of relationship between hepatic toxicity and acetylator phenotype in three thousand south Indian patients during treatment with isoniazid for tuberculosis. Am. Rev. Resp. Dis., 129, 58-61. Harvison, P.J., Guengerich, F.P., Rashed, M.S. & Nelson, S.D. (1988a). Cytochrome P-450 isozyme selectivity in the oxidation of acetaminophen. Chem. Res. Toxicol., 1, 47-52. Harvison, P.J., Egan, R.W., Gale, P.H., Christian, G.D., Hill, B.S. & Nelson, S.D. (1988b). Acetaminophen and analogs as cosubstrates and inhibitors of prostaglandin H synthase. Chem.-BioL Interactions, 64, 251-266.  105  Hein, D.W. & Weber, W.W. (1982). Polymorphic N-acetylation of phenelzine and monoacetylhydrazine by highly purified rabbit liver isoniazid Nacetyltransferase. Drug MetaboL Disp., 10, 225-229. Hinson, J.A. (1980). Biochemical toxicology of acetaminophen. Rev. Biochem. ToxicoL, 2, 103-129. Hinson, J.A., Pohl, L.R., Monks, T.J. & Gillette, J.R. (1981). Acetaminophen-induced hepatotoxicity. Life Sc!., 29, 107-116. Hoffman, J., Konopka, K., Buckhorn, C., Koop, D.R. & Waskell, L. (1989). Ethanolinducible cytochrome P450 in rabbits metabolizes enflurane. Br. J. Anaesth., 63, 103-108. Hofstra, A.H., Li-Muller, S.M. & Uetrecht, J.P. (1992). Metabolism of isoniazid by activated leukocytes: Possible role in drug-induced lupus. Drug MetaboL Disp., 20, 205-210. Holdiness, M.R. (1985). Chromatographic analysis of antituberculosis drugs in biological samples. J. Chromatogr., 340, 321-359. Hughes, H.B. (1953). On the metabolic fate of isoniazid. J. Pharmacol. Exp. Ther., 109, 444-452. Hyland, R., Gescher, A., Thummel, K., Schiller, C., Jheeta, P., Mynett, k., smith, A.W. & Mraz, J. (1992). Metabolic oxidation and toxication of N-methylformamide catalyzed by the cytochrome P450 isoenzyme CYP2E1. Molec. PharmacoL, 41, 259-266. Inaba, T., Tait, A., Nakano, M., Mahon, W.A. & Kalow, W. (1988). Metabolism of diazepam in vitro by human liver: independent variability of N-demethylation and C3-hydroxylation. Drug MetaboL Disp., 16, 605-608. Ingelman-Sundberg, M., Johansson, I., Persson, I., Hu, Y., Bergling, H., Ying, Q.Y., Bertilsson, L., Dahl, M.L. & Albano, E. (1992). Genetic polymorphism of ethanol-inducible cytochrome P4502E2. Relationship to function and incidence of liver cirrhosis (abstr.). PharmacoL & Toxicol., 71, Abstr. #19. Insel, P.A. (1990). Analgesic-antipyretic and antiinflammatory agents. In Goodman & Gilman's Pharmacological Basis of Therapeutics, eds.: Gilman, A.G., Rail, T.W., Nies, A.S. & Taylor, P., 8th ed. Pergamon Press, Toronto. pp. 656-659. lwainsky, H. (1988). Mode of action, biotransformation and pharmacokinetics of antituberculous drugs in animals and man. Handbook of Exp. PharmacoL, 84, 399-553. Jenner, P.J. & Ellard, G.A. (1987). Determination of the isoniazid metabolite monoacetylhydrazine in urine by high-performance liquid chromatography. J. Chromatogr., 415, 188-196. Jenner, P.J. & Ellard, G.A. (1989). Isoniazid-related hepatotoxicity: a study of the effect of rifampicin administration on the metabolism of acetylisoniazid in man. Tubercle, 70, 93-101.  106  Jo!low, D.J., Mitchell, J.R., Potter, W.Z. Davis, D.C., Gillette, J.R. & Brodie B.B. (1973). Acetaminophen-induced hepatic necrosis. II. Role of covalent binding in vivo. J. PharmacoL Exp. Ther., 187, 195-202. Jonville, A.P., Gauchez, A.S., Autret, E., Billard, C., Barbier, P., Nsabiyumra, F. & Breteau, M. (1991). Interaction between isoniazid and valproate: a case of valproate overdosage (lett.). Eur. J. Clin. PharmacoL, 40, 197-198. Juchau, M.R. & Horita, A. (1972). ^Metabolism of hydrazine derivatives of pharmacological interest. Drug MetaboL Rev., 1, 71-100. Kalow, W. (1987). Genetic variation in the human hepatic cytochrome P-450 system. Eur. J. Clin. Pharmacol., 31, 633-641. Kelley, J.M., Poet, R.B. & Chesner, L.M. (1952). Observations on the metabolic fate of nydrazid (isonicotinic acid hydrazide)(abstr.). Amer. Chem. Soc. 122nd Meeting, pp. 30. Kergueris, M.F., Bourin, M. & Larousse, C. (1986). Pharmacokinetics of isoniazid: influence of age. Eur. J. Clin. PharmacoL, 30, 335-340. Klos, C., Koob, M., Kramer, C. & Dekant, W. (1992). p-Aminophenol nephrotoxicity: biosynthesis of toxic glutathione conjugates. ToxicoL App!. Pharmacol., 115, 98-106. Koop, D.R. (1992). Oxidation and reductive metabolism by cytochrome P450 2E1. FASEB J., 6, 724-730. Krishnamurti, C.R. (1975). Isonicotinic acid hydrazide. In Antibiotics: Mechanism of Action of Antibacterial Agents, vol. 3, pp. 623-652. Eds.: Corcoran, J.W. & Hahn, F.E. Springer-Verlag, New York. Kronbach, T., Mathys, D., Umeno, M., Gonzalez, F.J. & Meyer, U.A. (1989). Oxidation of midazolam and triazolam by human liver cytochrome P450 II1A4. Molec. PharmacoL, 36, 89-96. Kubo, H., Kinoshita, T., Matsumoto, K. & Nishikawa, T. (1990). ^Fluometric determination of isoniazid and its metabolites in urine by high-performance liquid chromatography. Chromatographia, 30, 69-72. Kutt, H., Brennan, R., Dehejia, H. & Verebely, K. (1970). Diphenylhydantoin intoxication: a complication of isoniazid therapy. Am. Rev. Resp. Dis., 101, 377-384. Ladero, J.M., Jimenez, F.J., Benitez, J., Fernandez-Gundin, M.J., Martinez, C., Llerena, A., Cobaleda, J. & Munoz, J.J. (1989). Acetylator polymorphism in Parkinson's disease. Eur. J. Clin. PharmacoL, 37, 391-393. Lands, W.E.M. (1985). Mechanisms of action of antiinflammatory drugs. Adv. Drug. Res., 14, 147-164. Larrey, D., Letteron, P., Foliot, A., Descatoire, V., Degott, C., Geneve, J., Tinel, M. & Pessayre, D. (1986). Effects of pregnancy on the toxicity and metabolism of acetaminophen in mice. J. PharmacoL Exp. Ther., 237, 283-91.  107  Lauterburg, B.H., Smith, C.V. & Mitchell, J.R. (1981). Determination of isoniazid and its hydrazino metabolites, acetylisoniazid, acetylhydrazine, and diacetylhydrazine in human plasma by gas chromatography-mass spectrometry. J. Chromatogr., 224, 431-438. Lauterburg, B.H., Smith, C.V., Todd, E.L. & Mitchell, J.R. (1985a). Oxidation of hydrazine metabolites from isoniazid. Clin. PharmacoL Ther., 38, 566-571. Lauterburg, B.H., Smith, C.V., Todd, E.L. & Mitchell, J.R. (1985b). Pharmacokinetics of the toxic hydrazino metabolites formed from isoniazid in humans. J. Pharmacol. Exp. Ther., 235, 566-570. Lee, C.A., Thummel, K.E., Kalhorn, T.F., Nelson, S.D. & Slattery, J.T. (1991). Inhibition and activation of acetaminophen reactive metabolite formation by caffeine. Drug MetaboL Disp., 19, 348-353. Leibowitz, H., Krevolin, L. & Schwartz, A. (1989). Hemoperfusion treatment of massive isoniazid overdose. Delaware Med. J., 61, 71-73. Lewis, A. & Shepherd, R.G. (1970). Antimycobacterial Agents. In Medicinal Chemistry, pp. 409-491. Ed.: Burger, A. Wiley Interscience, New York. Lieber, C.S. (1988). Biochemical and molecular basis of alcohol-induced injury to liver and other tissues. New. EngL J. Med., 319, 1639-1650. Lind, R.C., Gandolfi, A.J. & Hall, P.M. (1990). Isoniazid potentiation of a guinea pig model of halothane-associated hepatotoxicity. J. AppL ToxicoL, 10, 161-165. Mandell, G.L. & Sande, m.A. (1990). Drugs used in the chemotherapy of tuberculosis and leprosy. In Goodman & Gilman's Pharmacological Basis of Therapeutics, eds.: Gilman, A.G., Rail, T.W., Nies, A.S. & Taylor, P., 8th ed. Pergamon Press, Toronto. pp. 1146-1159. Manyam, B.V., Ferraro, T.N. & Hare, T.A. (1987). Isoniazid-induced alteration of CSF neurotransmitter amino acids in Huntington's disease. Brain Res., 408, 125130. Mazze, R.I., Woodruff, R.E. & Heerdt, M.E. (1982). Isoniazid-induced enflurane defluorination in humans. Anesthesiology, 57, 5-8. McKennis, H., Yard, A.S., Weatherby, J.H. & Hagy, J.A. (1959). Acetylation of hydrazine and the formation of 1,2-diacetylhydrazine in vivo. J. Pharmacol. Exp. Ther., 126, 109-116. Merck Index (1989), 11th ed. Ed.: Budavari, S. Merck & Co., Rahway, NJ. pp. 8, 754, 816. Meyer, H. & MaIly, J. (1912). Uber hydrazinderivate der pyridincarbonsauren. Monatsh Chemie, 33, 393-414. Miners, J.0., Robson, R.A. & Birkett, D.J. (1986). Paracetamol metabolism in pregnancy. Br. J. Clin. PharmacoL, 22, 359-362.  108  Miners, J.0., Osborne, N.J., Tonkin, A.L. & Birkett, D.J. (1992). Perturbation of paracetamol urinary metabolic ratios by urine flow rate. Br. J. Clin. PharmacoL, 34, 359-362. Minton, N.A., Henry, J.A. & Frankel, R.J. (1988). Fatal paracetamol poisoning in an epileptic. Human ToxicoL, 7, 33-34. Mitchell, J.R. & Jo!low, D.J. (1975). Metabolic activation of drugs to toxic substances. GastroenteroL, 68, 392-410. Mitchell, J.R., Jollow, D.J., Potter, W.Z., Davis, D.C., Gillette, J.R. & Brodie, B.B. (1973a). Acetaminophen-induced hepatic necrosis. I. Role of drug metabolism. J. PharmacoL Exp. Ther., 187, 185-194. Mitchell, J.R., Jollow, D.J., Potter, W.Z., Gillette, J.R. & Brodie, B.B. (1973b). Acetaminophen-induced hepatic necrosis. IV. Protective role of glutathione. J. Pharmacol. Exp. Ther., 187, 211-217. Mitchell, J.R., Thorgeirsson, S.S., Potter, W.Z., Jo!low, D.J. & Keiser, H. (1974). Acetaminophen induced hepatic injury: protective role of glutathione in man and rationale for therapy. Clin. PharmacoL Ther., 16, 676-684. Mitchell, J.R., Long, M.W., Thorgiersson, U.P. & Jollow, D.J. (1975a). Acetylation rates and monthly liver function tests during one year of isoniazid preventive therapy. Chest, 68, 181-190. Mitchell, J.R., Thorgeirsson, U.P., Black, M., TimbreII, J.A., Snodgrass, W.R., Potter, W.Z., Jo!low, D.J. & Keiser, H.R. (1975b). Increased incidence of isoniazid hepatitis in rapid acetylators: possible relation to hydrazine metabolites. Clin. PharmacoL Ther., 18, 595-615. Mitchell, J.R., Zimmerman, H.J., Ishak, K.G., Thorgeirsson, U.P., TimbreII, J.A., snodgrass, W.R. & Nelson, S.D. (1976). lsoniazid liver injury: clinical spectrum, pathology, and probable pathogenesis. Ann. Intern. Med., 84, 181192. Moloney, S.J. & Prough, R.A. (1983). Biochemical toxicology of hydrazines. Rev. Biochem. Toxicol., 5, 313-348. Morgan, E.J., Koop, D.R. & Coon, M.J. (1983). Comparison of six rabbit liver cytochrome P-450 isozymes in formation of a reactive metabolite of acetaminophen. Biochem. Biophys. Res. Commun., 112, 8-13. Moulding, T.S., Redeker, A.G. & Kanel, G.C. (1991). Acetaminophen, isoniazid, and hepatic toxicity (lett.). Mn. Intern. Med., 114, 431. Muakkassah, S.F., Bidlack, W.R. & Yang, W.C.T. (1981). Mechanism of the inhibitory action of isoniazid on microsomal drug metabolism. Biochem. PharmacoL, 30, 1651-1658. Murphy, R., Swartz, R. & Watkins, P.B. (1990). Severe acetaminophen toxicity in a patient receiving isoniazid. Ann. Intern. Med., 113, 799-800. Erratum: Mn. Intern. Med., 114, 253 (1991).  109  Musch, E., Eichelbaum, M., Wang, J.k., von Sassen, W., Castro-Parra, M. & Dengler, H.J. (1982). Incidence of hepatotoxic side effects during antituberculous therapy (INH, RMP, EMB) in relation to the acetylator phenotype. Kiln. Wochenschr., 60, 513-519; Br. J. Clin. PharmacoL, 14, 575P-576P. Nebert, D.W. & Gonzalez, F.J. (1987). P450 genes: structure, evolution, and regulation. Ann. Rev. Biochem., 56, 945-993. Nelson, S.D. (1990). Molecular mechanisms of the hepatotoxicity caused by acetaminophen. Seminars in Liver Disease, 10, 267-278. Nelson, S.D. & Gordon, W.P. (1982). Metabolic fate of hydrazines. Adv. exp. Med. Biol., 136B, 971-981. Nelson, S.D., Mitchell, J.R., TimbreII, J.A., Snodgrass, W.R. & Corcoran, G.B. (1976). Isoniazid and iproniazid: activation of metabolites to toxic intermediates in man and rat. Science, 193, 901-903. Nelson, S.D., Dahlin, D.C., Rauckman, E.J. & Rosen, G.M. (1981). Peroxidasemediated formation of reactive metabolites of acetaminophen. Molec. PharmacoL, 20, 195-199. Nelson, S.D., Tirmenstein, M.A., Rashed, M.S. & Myers, T.G. (1991). Acetaminophen and protein thiol modification. Adv. Exp. Med. BioL, 283, 579-588. Noda, A., Goromaru, T., Matsuyama, K., Sogabe, K., Hsu, K. & lguchi, S. (1978). Quantitative determination of hydrazides derived from isoniazid in patients. J. Pharm. Dyn., 1, 132-141. Noda, A., Hsu, K.Y., Noda, H., Yamamoto, Y. & Kurozumi, J. (1983). Is isoniazidhepatotoxicity induced by the metabolite, hydrazine? J. UOEH, 5, 183-190. Noda, A., Sendo, T., Ohno, K., Hsu, K.Y., Yamamoto, Y., Matsuyama, K., Goto, S., Kurozumi, J. & Noda, h. (1984). Enhanced hepatitis of isoniazid induced by rifampicin coadministration. Role of the isoniazid metabolite, hydrazine. J. Pharm. Dyn., 7, s-81. Noda, A., Noda, H., Ohno, K., Sendo, T., Misaka, Y.A., Kanazawa, Y.R., Isobe, R. & Hirata, M. (1985). Spin trapping of a free radical intermediate formed during microsomal metabolism of hydrazine. Biochem. Biophys. Res. Commun., 133, 1086-1091. Noda, A., Ishizawa, M., Ohno, K., Sendo, T. & Noda, H. (1986). Relationship between oxidative metabolites of hydrazine and hydrazine-induced mutagenicity. Toxicol. Lett., 31, 131-137. Ochs, H.R., Greenblatt, D.J., Roberts, G.M. & Dengler, H.J. (1981). Diazepam interaction with antituberculosis drugs. Clin. PharmacoL Ther., 29, 671-678. Ochs, H.R., Greenblatt, D.J. & Knuchel, M. (1983). Differential effect of isoniazid on triazolam oxidation and oxazepam conjugation. Br. J. Clin. PharmacoL, 16, 743-746.  110  Offe, H.A., Siefken, W. & Domagk, G. (1952).^Hydrazinderivate aus pyridincarbonsauren mit carbonylrerbindungen und ihre wirksamkeit gegenuber Mycobacterium tuberculosis. Z. Naturforsch., 7B, 462-468. Parthasarathy, R., Sarma, R.G., Janardhanam, B., Ramachandran, P., Santha, T., Sivasubramaian, S., Somasundaram, P.R. & Tripathy, S.P. (1986). Hepatic toxicity in South Indian patients during treatment of tuberculosis with shortcourse regimens containing isoniazid, rifampicin and pyrazinamide. Tubercle, 67, 99-108. Penna, A. & Buchanan, N. (1991). Paracetamol poisoning in children and hepatotoxicity. Br. J. Clin. Pharmacol., 32, 143-149. Peretti, E., Karlaganis, G. & Lauterburg, B.H. (1987). Acetylation of acetylhydrazine, the toxic metabolite of isoniazid, in humans. Inhibition by concomitant administration of isoniazid. J. PharmacoL Exp. Ther., 243, 686-689. Perry, T.L., Kish, S.J., Hansen, S., Wright, J.M., Wall, R.A., Dunn, W.L. & Bellward, G.D. (1981). Elevation of brain GABA content by chronic low-dosage administration of hydrazine, a metabolite of isoniazid. J. Neurochem., 37, 3239. Perry, T.L., Wright, J.M., Hansen, S., Thomas, S.M.B., Allan, B.M., Baird, P.A. & Diewold, P.A. (1982) A double-blind clinical trial of isoniazid in Huntington disease. Neurology, 32, 354-358. Perry, T.L., Wall, R.A. & Hansen, S. (1985). Brain amino compounds in a Huntington's disease patient on isoniazid therapy. Neurology, 35, 755-758. Peters, J.H., Miller, K.S., & Brown, P. (1965a). The determination of isoniazid and its metabolites in human urine. Analytical Biochem., 12, 379-394. Peters, J.H., Miller, K.S. & Brown, P. (1965b). Studies on the metabolic basis for the genetically determined capacities for isoniazid inactivation in man. J. Pharmacol. Exp. Ther., 150, 298-304. Potter, W.Z., Davis, D.C., Mitchell, J.R., Jollow, D.J., Gillette, J.R. & Brodie, B.B. (1973). Acetaminophen-induced hepatic necrosis. III. Cytochrome P-450mediated covalent binding in vitro. J. PharmacoL Exp. Ther., 187, 203-210. Preece, N.E., Ghatineh, S. & Timbrell, J.A. (1990). course of ATP depletion in hydrazine hepatotoxicity. Arch. ToxicoL, 64, 49-53. Preece, N.E., Nicholson,. J.K. & Timbrell, J.A. (1991). Identification of novel hydrazine metabolites by ' "N-N M R. Biochem. PharmacoL, 41, 1319-1324. Preece, N.E., Forrow, S., Ghatineh, S., Langley, G.T. & Timbrell, J.A. (1992). Determination of hydrazine in biofluids by capillary gas chromatography with nitrogen-sensitive or mass spectrometric detection. J. Chromatogr., 573, 227234. Prescott, L.F. (1983). Paracetamol overdosage: pharmacological considerations and clinical management. Drugs, 25, 290-314.  111  Prescott, L.F., Mattison, P., Menzies, D.G. & Manson, L.M. (1990). The comparative effects of paracetamol and indomethacin on renal function in healthy female volunteers. Br. J. Clin. PharmacoL, 29, 403-412. Price-Evans, D.A. (1989). N-acetyltransferase. PharmacoL Ther., 42, 157-234. Prough, R.A., Freeman, P.C. & Hines, R.N. (1981). The oxidation of hydrazine derivatives catalyzed by the purified liver microsomal FAD-containing monooxygenase. J. Biol. Chem., 256, 4178-4184. Quemard, A., Lacave, C. & Laneelle, G. (1991). Isoniazid inhibition of mycolic acid synthesis by cell extracts of sensitive and resistant strains of Mycobacteriun aurum. Antimicrob. Agents. Chemother., 35, 1035-1039. Raucy, J.L., Lasker, J.M., Lieber, C.S. & Black, M. (1989). acetaminophen activation by human liver cytochromes P45011E1 and P450IA2. Arch. Biochem. Biophys., 271, 270-283. Rice, S.A., Maze, M., Smith, C.M., Kosek, J.C. & Mazze, R.I. (1987). Halothane hepatotoxicity in Fischer 344 rats pretreated with isoniazid. ToxicoL App!. PharmacoL, 87, 411-419. Rogers, S.M., Back, D.J., Stevenson, P.J., Grimmer, S.F.M. & Orme, M. (1987). Paracetamol interaction with oral contraceptive steroids: increased plasma concentrations of ethinyl estradiol. Br. J. Clin. PharmacoL, 23, 721-725. Rohrlich, M. (1951). Darstellung der nicotinursaure. Arch. Pharm., 284, 6-7. Rolband, G.C. & Marcuard, S.P. (1991).^cimetidine in the treatment of acetaminophen overdose. J. Clin. GastroenteroL, 13, 79-82. Ryan, D.E., Koop, D.R., Thomas, P.E., Coon, M.J. & Levin, W. (1986). Evidence that isoniazid and ethanol induce the same microsomal cytochrome P-450 in rat liver, an isozyme homologous to rabbit liver cytochrome P-450 isozyme 3a. Arch. Biochem. Biophys., 246, 633-644. Samigun, Mulyono & Santoso, B. (1990). Lowering of theophylline clearance by isoniazid in slow and rapid acetylators. Br. J. Clin. PharmacoL, 29, 570-573. Sanins, S.M., TimbreII, J.A., Elcombe, C., & Nicholson, J.K. (1992). Proton NMR spectroscopic studies on the metabolism and biochemical effects of hydrazine in vivo. Arch. ToxicoL, 66, 489-495. Sarkar, M.A., hunt, C., Guzelian, P.S. & Karnes, H.T. (1992). Characterization of human liver cytochromes P-450 involved in theophylline metabolism. Drug Metabol. Disp., 20, 31-37. Sarma, G.R., Kailasam, S., Nair, N.G.K., Narayana, A.S.L. & Tripathy, S.P. (1980). Effect of prednisolone and rifampin on isoniazid metabolism in slow and rapid inactivators of isoniazid. Antimicrob. Agents Chemother., 18, 661-666. Sarma, G.R., Immanuel, C., Kailasam, S., Narayana, A.S.L. & Venkatesan, P. (1986). Rifampin-induced release of hydrazine from isoniazid. Am. Rev. Resp. Dis., 133, 1072-1075.  112 Sato, C. & Lieber, C.S. (1981). Mechanism of preventive effect of ethanol on acetaminophen-induced hepatotoxicity. J. Pharmacol. Exp. Ther., 218, 811815. Seddon, C.E., Bobois, A.R. & Davies, D.S. (1987). Comparative activation of paracetamol in the rat, mouse and man. Arch. Toxicol., Suppl. 11, 305-309. Seeff, L.B., Cuccherini, B.A., Zimmerman, H.J., Adler, E. & Benjamin, S.B. (1986). Acetaminophen hepatotoxicity in alcoholics. Ann. Intern. Med., 104, 399-404. Sendo, T., Noda, A., Noda, H., Hsu, K.Y. & Yamamoto, Y. (1984). Metabolic hydrolysis of isoniazid by subcellular fractions of rat liver. J. UOEH, 6, 249-255. Shoeb, FLA., Bowman, B.U., Ottolenghi, A.C. & Merola, A.J. (1985). Peroxidase mediated oxidation of isoniazid. Antimicrob. Agents Chemother., 27, 399-412. Sinha, B.K. (1983). Enzymatic activation of hydrazine derivatives. J. Biol. Chem., 258, 796-801. Slattery, J.T., Wilson, J.M., Kalhorn, T.F. & Nelson, S.D. (1987). Dose-dependent pharmacokinetics of acetaminophen: evidence of glutathione depletion in humans. Clin. Pharmacol. Ther., 41, 413-418. Smith, J.A.E., Hine, I.D., Beck, P. & Routledge, P.A. (1986). Human Toxicol., 5, 383385. Snodgrass, W., Potter, W.Z., TimbreII, J., Jollow, D.J. & Mitchell, J.R. (1974). possible mechanisms of isoniazid related hepatic injury (abstr.). Clinical Res., 22, 323A. Sokal, R.R. & Rohlf, F.J. (1987). Introduction to Biostatistics. 2nd ed. pp. 204-207. W.H. Freeman & Co., New York. Speck, R.F. & Lauterburg, B.H. (1991). Fish oil protects mice against acetaminophen hepatotoxicity in vivo. Hepatology, 13, 557-561. Spika, J.S. & Aranda, J.V. (1987). Interaction between chloramphenicol and acetaminophen. Arch. Dis. Child., 62, 1087-1088. Springer, D.L., Krivak, B.M., Broderick, D.J., Reed, D.J. & Dost, F.N. (1981). Metabolic fate of hydrazine. J. Toxicol. Environ. Health, 10, 941-953. Steele, M.A., Burk, R.F. & DePrez, R.M. (1991). Toxic hepatitis with isoniazid and rifampin: a meta-analysis. Chest, 99, 465-471. Stein, C.M., Thornhill, D.P., Neill, P. & Nyazema, N.Z. (1989). Lack of effect of paracetamol on the pharmacokinetics of chloramphenicol. Br. J. Clin. Pharmacol., 27, 262-264. Stevens, J.L. & Jones, D.P. (1989). The mercapturic acid pathway: biosynthesis, intermediary metabolism, and physiological disposition. In Glutathione: Chemical, Biochemical, and Metabolic Aspects, Part B, pp. 45-84. Eds.: D. Dolphin, 0. Avramovic, and R. Poulson. Wiley & Sons, Toronto.  113  Thomas, b.H., Zeitz, W. & Whitehouse, L.W. (1987). Effect of rifampin, phenobarbital pretreatment, and acetylator phenotype on acetylisoniazid metabolism in the rabbit. Can. J. Physiol. PharmacoL, 65, 419-423. TimbreII, J.A. (1979).^Studies on the role of acetylhydrazine in isoniazid hepatotoxicity. Arch. ToxicoL, Suppl. 2, 1-8. TimbreII, J.A. & Wright, J.M. (1979). Studies on the effect of isoniazid on acetylhydrazine metabolism in vivo and in vitro. Drug Metabol. Disp., 7, 237240. TimbreII, J.A., Wright, J.M. & Baillie, T.A. (1977a). Monoacetylhydrazine as a metabolite of isoniazid in man. Clin. Pharmacol. Ther., 22, 602-608. TimbreII, J.A., Wright, J.M. & Smith, C.M. (1977b). Determination of hydrazine metabolites of isoniazid in human urine by gas chromatography. J. Chromatogr., 138, 165-172. TimbreII, J.A., Mitchell, J.R., Snodgrass, W.R. & Nelson, S.D. (1980). Isoniazid hepatotoxicity: the relationship between covalent binding and metabolism in vivo. J. PharmacoL Exp. Ther., 213, 364-369. TimbreII, J.A., Scales, M.D.C. & Streeter, A.J. (1982). studies on hydrazine hepatotoxicity. 2. Biochemical findings. J. ToxicoL Environ. Health, 10, 955968. TimbreII, J.A., Park, B.K. & Harland, S.J. (1985). A study of the effect of rifampicin on isoniazid metabolism in human volunteer subjects. Human ToxicoL, 4, 279285. van de Straat, R., Vromans, R.M., Bosman, P., de Vries, J. & Vermeulen, N.P.E. (1988). Cytochrome P-450-mediated oxidation of substrates by electrontransfer; role of oxygen radicals and of 1- and 2-electron oxidation of paracetamol. Chem.-Biol. Interactions, 64, 267-280. Vermeulen, N.P.E., Bessems, J.G.M. & van de Straat, R. (1992). Molecular aspects of paracetamol-induced hepatotoxicity and its mechanism-based prevention. Drug MetaboL Rev., 24, 367-407. von Sassen, W., Castro-Parra, M., Musch, E. & Eichelbaum, M. (1985). Determination of isoniazid, acetylisoniazid, acetylhydrazine and diacetylhydrazine in biological fluids by hligh-performance liquid chromatography. J. Chromatogr., 338, 113-122. Weber, W.W. (1987). The Acetylator Genes and Drug Response. Oxford University Press, New York. Weber, W.W. & Hein, D.W. (1979). Clinical pharmacokinetics of isoniazid. Clin. Pharmacokin., 4, 401-422. Weber, W.W. & King, C.M. (1981). N-acetyltransferase and arylhydroxamic acid transferase. Methods in EnzymoL, 77, 272-280. Weber, W.W., Radtke, H.E. & Tanner, R.H. (1980). Extrahepatic N-acetyltransferases and N-deacetylases. In: Extrahepatic Metabolism of Drugs and Other Foreign  114  Compounds, pp. 493-531. Ed.: T.E. Gram. SP Medical & Scientific Books, New York. Whitehouse, L.W., Tryphonas, L., Paul, C.J., Solomonraj, G., Thomas, B.H. & Wong, L.T. (1983). Isoniazid-induced hepatic steatosis in rabbits; an explanation for susceptibility and its antagonism by pyridoxine hydrochloride. Can. J. PhysioL PharmacoL, 61, 478-487. Wilkinson, L. (1990a). SYGRAPH: The System for Graphics. Evanston, Ill.: SYSTAT Inc. pp. 165-169. Wilkinson, L. (1990b). SYSTAT: The System for Statistics. Evanston, Ill.: SYSTAT Inc. pp. 263-264. Woodward, K.N. & TimbreII, J.A. (1984). Acetylhydrazine hepatotoxicity: the role of covalent binding. Toxicology, 30, 65-74. Woo, J., Chan, C.H.S., Walubo, A. & Chan, K.K.C. (1992). Case report: Hydrazine a possible cause of isoniazid-induced hepatic necrosis. J. Med., 23, 51-59. Wright, J.M. (1992). Drug interactions. In Melmon & Morrelli's Clinical Pharmacology: Basic Principles in Therapeutics, 3rd ed. Eds.: Melmon, K.W., Morrelli, H.F., Hoffman, B.B. & Nierenberg, D.W. McGraw Hill, Toronto. pp. 1012-1021. Wright,i q.m. [1  & Timbrell, J.A. (1978).^Factors affecting the metabolism of ^in rats. Drug Metabol. Disp., 6, 561-566.  Wright, J.M. & Wall, R.A. (1985). Isoniazid hepatotoxicity is dose related in man (abstr.). 4th South-East Asian/ Western Pacific Regional Meeting of Pharmacologists, 7-11 May, Penang, Malaysia. Wright, J.M., Stokes, E.F. & Sweeney, V.P. (1982). lsoniazid-induced carbamazepine toxicity and vice versa: a double drug interaction. New. EngL J. Med., 307, 1325-1327. Wright, J.M., Wall, R.A. & Adams, S.P. (1984). Carbamazepine-induced isoniazid hepatotoxicity: a serious drug interaction (abstr.). Clin. Invest. Med., 7 (Suppl. 2). Wright, J.M., Ngai, H., Adams, S., Bain, A. & Wall, R.A. (1986). Lack of hepatotoxicity of acetylhydrazine in rodents (abstr.). Acta PharmacoL ToxicoL, 59 (Suppl. 5), 221. Zamboni, V. & Defranceschi, A. (1954). Identification of isonicotinoylhydrazones of pyruvic and a-ketoglutaric acid in rat urine after treatment with isonicotinic acid hydrazide (isoniazid). Biochim. Biophys. Acta, 14, 430-432. Zand, R., Nelson, S.D., Slattery, J.T., Thummel, K.E., Kalhorn, T.F., Adams, S.P. & Wright, J.M. (1992). Inhibition and induction of cytochrome P450 2E1catalyzed oxidations by isoniazid in humans. (Manuscript in preparation). Zar, J.H. (1984). Biostatistical Analysis, 2nd ed. Prentice-Hall, Englewood Cliffs, NJ.  115  Zhang, Y., Heym, B., Allen, B., Young, D. & Cole, S. (1992). The catalase-peroxidase gene and isoniazid resistance of Mycobacterium tuberculosis. Nature, 358, 591-593.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086189/manifest

Comment

Related Items