- Email: [email protected]
Physiological factors affecting drug toxicity
REGULATORY
TOXICOLOGY
AND
PHARMACOLOGY
Physiological
of Drug
(1983)
Factors Affecting KEITH
Bureau
3, 389-398
Research, Ottawa,
Health Ontario
Received
Drug Toxicity’
BAILEY Protection Branch, KlA OL2. Canada
TunneyS
Pasture,
Juiy 6. 1983
Physiological factors that affect the fate of drugs in the body and thereby have effects on their pharmacology and toxicology involve the systemsthat control absorption, distribution, metabolism, and excretion. The main factors are disease, genetics, and age. Nutritional status, sex, hormonal status (e.g., the effectsof pregnancy), and circadian rhythm have important influences. Maternal toxicity will affect the fetus. The absorption and excretion of drugs are frequently reduced by diseases. Excretion is reduced by impaired renal function, often found in the elderly, which may increase the toxic response. Distribution is affected by body weight and build, for example, the proportion of fat. The disposition of many drugs has been shown to be significantly influenced by circadian rhythms such that therapeutic doses may exhibit toxicity if administered at an inappropriate time of day. Metabolism is modified by environmental influences including previous food and drug experience, and various factors such as age, sex, and disease. Intersubject variations in drug disposition can be very great with possibly severe consequences for the individual; in this regard, knowledge of genetic polymorphism in drug metabolizing enzymes is rapidly increasing. The toxicology of a drug may be profoundly affected by a particular disease state, for example, the administration of a drug that might be a tumor promoter when a cancerous or precancerous condition exists. These effectsare illustrated with examples from the literature and recent studies undertaken in the Bureau of Drug Research.
Physiological factors that affect drug toxicity are multiple, interrelated in complex ways, and therefore difficult to unravel especially in human subjects. In this brief review of a very large topic it is intended to illustrate some of the effects that are known using examples from the literature and studies recently undertaken in the Bureau of Drug Research (BDR). AGE As yet, there are no studies that have followed a drug’s disposition in an individual from childhood into youth and adulthood, or from adulthood into old age. However, it is apparent that the clinical pharmacokinetics of many drugs are different in the newborn, the child, the adult, and the elderly. The physiological characteristics of ’ Adapted from a presentation at the February 21-23, 1983, meeting of the Toxicology Forum, Arlington, Va. 389
0273-2300183
$3.00
390
KEITH
BAILEY
TABLE
PHYSIOLOGICAL FACTORSCHARACTERISTIC
1
OF THE OLD
THAT AFFECT DRUG TOXICITY~
Absorption from GI tract possibly reduced Higher proportion of fat Reduced plasma protein binding Prolongation of plasma drug half-life (reduced metabolism?) Reduced renal elimination Multiple diseases Nutritional deficiencies Altered tissue response (receptor modification?) ’ Crooks (1976) Vestal (1978), Ricbey and Bender (1977).
aging listed in Table 1 influence drug disposition and have pharmacological consequences. Pathophysiological features of the very young (Table 2) and of infants and children and the consequent effects on drug efficacy and toxicity have been reviewed (Table 2). Morselli et al. (1980) presented detailed pharmacokinetic profiles of two dozen drugs that are commonly prescribed for children. An interesting relationship exists between theophylline and caffeine. Theophylline is used for the treatment of asthma in children and adults and for apnea and bradycardia in premature infants. In premature newborns it may be metabolized to caffeine which, like theophylline, is effective in recurrent apnea. Caffeine has an apparent plasma half-life of 43-231 hr compared with 12-64 hr for theophylline. It has been suggested (Morselli et al., 1980) that the concomitant presence of caffeine in premature infants could explain the lowered threshold for toxic effects observed with theophylline. The second example is the use of chloramphenicol in infants a few days old, where the half-life is about 26 hr in contrast to 4 hr in older children. This is now known to be due to the inability of neonates to glucuronidate the parent drug, but before this was realized, persistent high blood concentrations led to the deaths of infants who received a child’s dose. GENETICS The overall disposition of a drug depends on several processes that are under genetic control and possible toxic consequences are difficult to predict. Comparisons of the TABLE 2 PHYSIOLOGICALFACTORS CHARACTERISTIC OF THE VERY YOUNG THAT AFFECT DRUG TOXICITY a Gastric emptying prolonged Gastric pH neutral Higher proportion of body water Increased percutaneous absorption Reduced plasma protein binding Lowered hepatic microsomal activity Ready stimulation of metabolic capability (liver greater proportion of body mass) Lowered glucuronidation Reduced renal capability u Rylance (I 98 I), Morselli et at. (1980) Yaffe and Juchau (1974).
FACTORS
AFFECTING
391
DRUG TOXICITY
metabolism of drugs in populations that belong to different races or ethnic groups have revealed distributions of populations that handle drugs in distinctive ways. Table 3 shows the distribution of fast and slow acetylators of the antitubercular drug isoniazid (INH). Isoniazid has been extensively used among the autochthonic populations of North America alone in prophylaxis or in combination with other tuberculostats in treatment programs. Neuropathic toxic symptoms are more common with slow acetylators and signs of liver toxicity have been associated with fast acetylators; this point is somewhat controversial and it has been pointed out that homozygous and heterozygous rapid acetylators may differ in their susceptibilities to adverse effects (Kalow, 1982). Enzymes other than acetyl transferases are involved in INH toxicity. Thus, previous studies have attributed cell necrosis to the oxidative activation of the acetyl moiety of acetylhydrazine (Nelson et al., 1976; Timbre11 et al., 1980), and bisp-nitrophenyl phosphate is an amidase inhibitor that decreases the extent of hepatic necrosis induced by INH and acetyl-INH (Mitchell et al., 1976). In recent BDR studies, the specificity of hepatic microsomal amidase for the amide bonds found in INH and acetyl-INH was examined in male rats, guinea pigs, and rabbits of intermediate acetylator status. Rabbits were found to possess a metabolic profile reasonably similar to that of man. The rabbit showed the greatest hydrolytic activity for the amide bonds (Table 4). In each species the isonicotinoyl bond of acetyl-INH was the most susceptible. Furthermore, the rabbit amidase had a fourfold greater affinity for acetyl-INH than for INH (Michaelis constant data). Covalent binding studies with [‘4C]acetylhydrazine and [‘4C]acetyl-INH in rats and rabbits showed that there was less binding of the former, but more binding of [‘4C]acetylINH in the rabbit. Also, in the rabbit a greater proportion of the acetylhydrazine is metabolized to diacetylhydrazine. In rabbits, the most susceptible of the three species to INH hepatotoxicity, it seems to be the high amidase activity that produces higher amounts of acetylhydrazine and causes the high level of binding. Human amidase studies would therefore be very interesting. A well-known example of genetic control is the benzylic oxidation of &hisoquin investigated by Smith’s group in London (Price-Evans et al., 1980). Approximately 10% of the British population are homozygous for poor metabolizing capacity, 50% are homozygous rapid and 40% are heterozygous rapid metabolizers (Kalow, 1981). There was a 40-fold difference in individual dose requirements for a given therapeutic (antihypertensive) effect. TABLE 3 DISTRIBUTION(%)•
FACETYLATOR.SOFISONIAZID"
Fast Population
Slow
Heterozygous
Homozygous
Indian Caucasian Negro Eskimo Japanese Chinese
59 59 55 II 12 22
36 36 39 44 45 50
5 5 7 46 43 28
a See Kalow (1982).
KEITH
392
BAILEY
TABLE 4 HYDROLYTIC ACWITIES OFHEPATICMICROSOMAL PROTEIN'
Py *CO * NHNH-COCH, Py - CO-NHNH . COCH3 Py - CO-NI-INH2
Rat
Guinea pig
Rabbit
0.7 4.0 1.6
4.7 5.8 0.3
6.0 153 30
LIUnits are nmol/hr/mg protein (Thomas and Whitehouse, personal communication).
A recent report by Lennard et al. (1982) shows that the Pi-adrenoceptor antagonist metoprolol exhibits genetic polymorphism of the debrisoquin type. Plasma metoprolol concentrations were much higher in the poor hydroxylators, and this group had six times the area under the plasma concentration vs time curve (AUC) and three times the elimination half-life. Beta blockade was well maintained 24 hr after treatment in poor hydroxylators but was negligible in extensive hydroxylators. These workers suggested that asthmatic patients who are poor hydroxylators might be at some risk if treated with standard doses of metoprolol.2 Other associations with debrisoquin hydroxylation deficiency are shown in Table 5. It can be inferred that higher cancer rates are associated with a greater oxidative ability to produce active, toxic metabolites from various xenobiotics. There is scope for prospective studies along these lines. As part of BDR’s bioavailability assessment of quinidine formulations, McGilveray et al. ( 198 1) investigated pharmacokinetic variations in humans. Intersubject variation in the AUC was found to be high. The extremes of terminal half-lives among 24 subjects were 9.49 and 6.24 hr, the plasma quinidine concentrations at 24 hr differing by a factor of 10. In the BDR study, the frequency distribution of clearance appeared to be normal. The plasma binding of disopyramide (DP) and mono-Wdealklyldisopyramide (MND) have been reported recently by Bredesen et al. (1982). The drug and its metabolite have antiarrhythmic effects. Plasma from 70 patients on maintenance therapy but otherwise drug free was analyzed for the two substances. A wide variation was found: from 1 to 20 ctmol/liter for DP and from 1 to 16 ~mol/liter for MND and MND:DP ratios of 1.6-4.0. Partial data from the plasma of healthy subjects are shown in Table 6. It was found that the unbound fraction of both DP and MND increased with higher concentrations and, further, that the unbound fraction of each one was increased in the presence of clinically important concentrations of the other. According to these authors, the interpatient variance of unbound DP concentration might be tenfold or higher in the assumed therapeutic range. Unbound drug and metabolite concentrations should be monitored in order to minimize toxic effects and maximize the therapeutic benefit. The use of drugs with such characteristics in patients with cardiac disease is complicated by the affect of the disease itself on disposition. * This matter has received further attention very recently; see D. B. Jack, M. J. Kendall, M. Wilkins, and C. P. Quarterman (1983). N. Engl. J. Med. 308,964; P. Dayer, F. Courvoisier, L. Balant, and J. Fabre (1983). N. Engl. J. Med. 3B8, 964-965; M. S. Lennard, S. Freestone, L. E. Ramsay, G. T. Tucker, H. F. Woods, and J. H. Silas (1983). N. Engl. J. Med. 308, 965.
FACTORS
AFFECTING
393
DRUG TOXICITY
TABLE 5 DECREASED
DEBRWQUINE
METABOLISM
Decreased metabolism associations Sparteine Guanoxan Phenacetin Perhexiline Nortriptyline Phenformin Propranolol Metoprolol Phenytoin Mephenytoin
AWXIATIONS”
Normal metabolism associations Amobarbital Antipyrine Tolbutamide Clozapine Possible decrease associations Liver cancer (? aflatoxin) Bronchial carcinoma
a cf. Kalow (198 I).
DIETARY
CONSIDERATIONS
Dietary habits affect body weight (proportion of fat, enzyme levels, etc.), gut pH, flora, transit times, and so forth, and so considerably modify pharmacological responses (Fig. 1). Thus, investigations into the hepatotoxicity of CC& indicated that greater toxicity arose among animals receiving a lower proportion of carbohydrate but increased fat and protein in isocaloric diets (Nakajima et al., 1982). On a superficial level, if a drug is taken with meals the outcome can be very different from that following a period of starvation. For example, propranolol qid followed by meals produced maximum plasma levels and AUCs that were twice those in fasting subjects. It was suggested that food decreases presystemic elimination by increasing the rate of blood flow in the viscera. (Dvomik et al., 1982). During investigations into the bioavailability of metronidazole formulations, McGilveray et al. (1978) noted that a reference solution gave a lower bioavailability in starved subjects than did a reference tablet. It was suggested that in some individuals, gastrointestinal flow was so rapid TABLE 6 DISOPYRAMIDE
(DP)
AND
MONODEALKYLDISOPYRAMIDE FROM FIVE HEALTHY SUBJECTS”
(MND)
Mean equilibrium (mol/U DP initial (wmol/l) 1
5 10 15 20 ’ The assumed therapeutic
0.8 4.0 7.0 10.6 12.8 3.8 12.4 range
Mean
0 0 0 0 0 14.4 14.3 of DP
is 6-15
PLASMA
DP unbound fraction (‘%)
MND
DP
IN SPIKED
rmol/l
+ SD
12* 19f 28k 34+ 43i 35 Ik 54k (Bredesen
et al.,
2 5 4 5 8 10 10 1982).
Range 10-14 14-28 22-3 1 29-41 34-50 25-46 40-64
394
KEITH
BAILEY
FOOD
GENERAL
DRUG-PROTEIN
NUTRUTION
PH
ACCUMULATION OF LIPOPBILIC DRUGS I ACIDIC/BASIC DRUG ABSORPTION, ELIMINATION
BINDING I ENZYME LEVELS, BIOTRANSFORMATION, ELIMINATION
MOBILIZATION
MOBILIZATION -.l\/--MODIFIED
EFFECTS
FIG.1. Physiological associations of food intake with drug pharmacology and toxicology.
that solutions rapidly passed the absorption sites. The disintegrating tablets moved more slowly in the area of absorption, with particles being held in the mucosal layer. Feldman et al. (1982) showed that theophylline kinetics in children are affected by the diet. The half-life was 6.8 hr on a normal diet, 4.8 hr on a high protein diet, and 18.1 hr on a high carbohydrate diet. Significant changes in serum levels could occur in l-2 weeks. PHYSIQUE From diet we make a natural progression to physique. A curious example was provided recently of an apparent sex difference in dapsone treatment of leprosy in Ethiopian men and women (Modderman et al., 1982). These patients received 900mg injections into the buttocks using a 3.5-cm-long needle. Serum concentrations reached 2 pg/ml on Day 3 in men, declining to about 0.5 pg by Day 10 and declining steadily to 0.1 pg/ml by Day 28. In women, the concentration was 1 &ml on Day 3, declining steadily to 0.4 pg/ml by Day 28. Almost certainly, the women received the injections into the subcutaneous fat and the men received injections into the muscle. Cockshott and Thompson (1982) pointed out that the sexual dimorphism in total fat content (women 24%, men 17%) extends to its distribution. Pregnancy is a condition that as yet remains sex specific! One of its effects is to increase glomerular filtration rate by 30-50% and renal plasma flow by about 25%, values returning to normal soon after delivery (Dvorchik, 1982). Therefore, dosages of drugs that are cleared by renal mechanisms may need to be adjusted during and after pregnancy if toxic effects are to be avoided. An instance of lithium toxicity due to failure to make a dose adjustment postpartum has been attributed to these changes.
FACTORS
AFFECTING
395
DRUG TOXICITY
Parsons and Pelletier (1982) suggested that inhibition of the hepatic mixed-function oxidase system could be responsible for the delayed elimination of caffeine that was observed near term, where levels may be three times normal. The effects of posture on pharmacokinetic parameters of drugs have been studied by several groups. This simple physiological variable can also change the toxicological profile. A recent report by Heller et al. (1982) associated the prior consumption of nonsteroidal anti-inflammatory drugs (NSAIs) with the development of benign esophageal stricture, a consequence of simple mechanics. NSAIs have long been associated with gastric erosions. Hey et al. (1982) investigated the esophageal transit of common types of tablets and capsules. They subsequently recommended that subjects should swallow the preparation with at least 100 ml of water and remain standing for at least 90 sec. CIRCADIAN
RHYTHMS
It seems to be popular for people to blame a maliferous phasing of their biorhythms when everything seems to go wrong at once. Circadian pharmacokinetics have been demonstrated for the drugs and related xenobiotics shown in Table 7. Cisplatin provides a particularly interesting example. This very useful but unfortunately poorly tolerated anticancer drug shows the lowest nephrotoxicity when given in the evening, which is when the urinary concentration is lowest. The highest antitumor activity is simultaneous with best tolerance. Circadian rhythms provide yet another complicating factor in evaluating toxic responses. DISEASE Although disease is a pathological state it is appropriate to consider it in the context of physiological factors. There are numerous toxicological effects to be expected when drugs are administered to a diseased patient in comparison with results from healthy subjects in whom drugs are tested. In general, the metabolic clearance may be reduced if liver function is impaired and thus total plasma clearance reduced even when the renal clearance is normal. Drug responses are increased when there is renal failure since excretion is reduced. The plasma levels of drugs with high liver extraction ratios may be increased when patients are hypotensive. TABLE 7 XENOBIOTICS
Acetaminophen Aspirin Digitalis Hexobarbital Nortriptyline Potassium chloride Sulfasymazine Theophylline
EXHIBITING
CHRONOPHARMACOKINETICS’
Aminopyrine Cisplatin Erythromycin Indomethacin Phenacetin Sodium salicylate Sulfonilamide Xylose
Ampicillin Clorazepate Ferrous sulfate Lithium carbonate Phenytoin Soman Sulindac
a Reinberg and Smolensky (1982), Elsmore (198 I), Reinberg and Halberg (197 1).
396
KEITH
BAILEY
TABLE 8 IN VITRO GENOTOXKITY TESTINGOFDIAZEPAM~ 1. Diazepam did not induce DNA repair in freshly isolated rat hepatocytes in subtoxic dose range. 2. Diazepam did not alfect DNA repair induced by methylmethane sulfonate in freshly isolated rat hepatocytes. 3. Mutations by 8-azaguanine at HGPRT locus in rat liver cell line T5lB were not induced by diazepam, with or without metabolic activation. 4. Diazepam did not induce DNA synthesis in quiescent cells (low Ca). b 5. Diazepam and oxazepam, with or without metabolic activation, were unable to induce reverse mutations in Salmonella typhimurium strains TA 98 and TA 100. 6. As in 5 (above) except with a histidine-enriched
medium (Batzinger et al., 1978).
’ Matula and Swierenga (personal communication). b Boynton and Whitfield (1980).
Studies recently carried out in our laboratories on diazepam represent a special example of the effect of disease on possible drug toxicity. Diazepam is one of the most commonly prescribed tranquilizers, and a great deal of controversy arose following some suggestions (Horrobin et al., 1979; Hoi-robin, 198 1) that it might be a promoter of breast cancer in women. No evidence has accrued from epidemiological studies that diazepam is a tumor promoter. We undertook in vitro studies designed to determine genotoxic activity, with the results shown in Table 8. (The stimulation of DNA synthesis in calcium-deprived cells is a short-term promoter assay described by Boynton and Whitfield (1980) Canadian National Research Council.) In the meantime, in vivo studies in female rats induced for mammary tumors by the administration of dimethylbenzanthracene were started according to the protocol shown in Table 9. This pilot study provided no evidence for the promotion of palpable tumors. This last example is a rather particular one of drug toxicity that could be sex and disease related (and because of the type of disease, also age related). TABLE 9 SUMMARYOFDIAZEPAM PILOTSTUDYINFEMALERATS" Virgin female Sprague-Dawley rats purchased at 25 days 54 days old: 5 mg dimethylbcnzanthracene in 1.0 ml corn oil, gavage Group 1 (control, 19): vehicle (gum tragacanth) Group 2 (9): diazepam 1.0 mg/kg/day, SX/week Group 3 (8): diazepam 10.0 mg/kg/day, SX/week Mammary regions palpated weekly All surviving animals killed after 45 weeks and autopsied Employ life table analysis to diminution of prolongation of time to tumor. u Matula (personal communication).
FACTORS
AFFECTING
DRUG
397
TOXICITY
ACKNOWLEDGMENTS The comments and advice of G. L. Frederick, T. Matula, Thomas, and L. W. Whitehouse are greatly appreciated.
I. J. McGilveray,
S. H. H. Swierenga,
B. H.
REFERENCES BATZINGER, R. P., BUEDING, E., REDDY, B. S., AND WEISBURG, J. H. (1978). Formation of a mutagenic drug metabolite by intestinal microorganisms. Cancer Res. 38, 608-6 12. BOYNTON, A. L., AND WHITRELD, J. F. (1980). Stimulation of DNA synthesis in calcium-deprived T5 1B liver cells by the tumour promoters phenobarbital, saccharin, and 12-O-tetradecanoylphorbol13-acetate. Cancer Res. 40,4541-4545. BREDESEN, J. E., PIKE, E., AND LUNDE, P. K. M. (1982). Plasma binding of disopyramide and mono-Ndealkyldisopyramide. Brit. J. Clin. Pharmacol. 14, 673-676. COCKSHOTT, W. P., AND THOMPSON, G. T. (1982). Injections into fat instead of muscle. N. Engl. J. Med. 307, 1581. CROOKS, J., O’MALLEY, K., AND STEVENSON, I. H. (1976). Pharmacokinetics in the elderly. Clin. Pharmacokinet. 1, 280-296. DVORCHIK, B. H. (1982). Drug disposition during pregnancy. Zntl. J. Biol. 3, 129-137. DVORNIK, D., KRAML, M., DUBUC, J., FENCIK, M., WEIDLER, D., AND MULLANE, J. F. (1982). Propranolol concentrations in healthy men given 80 mg daily in divided doses: Effect of food and circadian variation. Curr. Ther. Res. 32, 214-224. ELSMORE, T. F. (198 1). Circadian susceptibility to soman poisoning. Fundam. Appl. Toxicol. 1.238-24 1. FELDMAN, C. H., HUTCHINSON, V. E., SHER, T. H., FELDMAN, B. R., AND DAVIS, W. J. (1982). Interaction between nutrition and theophylline metabolism in children. Ther. Drug Monit. 4,69-76. HELLER, S. R., FELLOWS, 1. W., OGILVIE, A. L., ANLI ATKINSON, M. (1982). Non-steroidal anti-inflammatory drugs and benign oesophageal stricture. Brit. Med. J. 285, 167-168. HEY, H., JORGENSEN, F., WRENSEN, K., HASSELBACH, H., AND WAMBERG, T. (1982). Oesophageal transit of six commonly used tablets and capsules. Brit. Med. J. 285, 17 17- 17 19. HORROBIN, D. F., GHAYUR, T., AND KARMALI, R. A. (1979). Mind and cancer. Luncet 1, 978. HORROBIN, D. F. (1981). Diazepam as tumour promoter. Lancef 1, 277-278. KALOW, W. (198 1). The metabolism of xenobiotics in different populations. Canad. J. Physiol. Pharmacol. 60, l-12. KALOW, W. (1982). Ethnic differences in drug metabolism. Clin. Pharmacokinet. 7, 373-400. LENNARD, M. S., SILAS, J. H., FREESTONE, S., RAMSAY, L. E., TUCKER, G. T., AND WOODS, H. F. (1982). Oxidation phenotype-A major determinant of metoprolol metabolism and response. N. Engl. J. Med. 307, 1558-1560. MCGILVERAY, 1. J., MIDHA, K. K., Loo, J. C. K., AND COOPER, J. K. (1978). Bioavailability of commercial metronidazole formulations. Intl. J. Clin. Pharmacol. 16, 110-I 15. MCGILVERAY, I. J., MIDHA, K. K., ROWE, M., BEAUDOIN, N., AND CHARETTE, C. (1981). Bioavailability of 11 quinidine formulations and pharmacokinetic variations in Humans. J. Pharm. Sci. 70, 524-529. MITCHELL, J. R., ZIMMERMAN, H. J., ISHAK, K. G., THORGEIRSSON, U. P., TIMBRELL, J. A., SNODGRASS, W. R., AND NELSON, S. D. (1976). Isoniazid liver injury: Clinical spectrum, pathology, and probable pathogenesis. Ann. Intern. Med. 84, 18 l-192. MODDERMAN, E. S. M., HILBERS, H. W., ZUIDEMA, J., MERKUS, F. W. H. M., AND WARNDORFF, T. (1982). Injections into fat instead of muscle. N. Engl. J. Med. 307, 1581. MORSELLI, P. L., FRANC~MORSELLI, R., AND Bossr, L. (1980). Clinical pharmacokinetics in newborns and infants. Clin. Pharmacokinet. 5, 485-527. NAKAJIMA, T., KOYAMA, Y.. AND SATO, A. (1982). Dietary modification of metabolism and toxicity of chemical substances-With special reference to carbohydrate. B&hem. Pharmacol. 31, 1005-1011. NELSON, S. D., MITCHELL, J. R., TIMBRELL, J. A., SNODGRASS, W. R., AND CORCORAN, G. B. (1976). Isoniazid and iproniazid: Activation of metabolites to toxic intermediates in man and rat. Science 193, 901-903. PARSONS, W. D., AND PELLETIER, J. G. (1982). Delayed elimination of caffeine by women in the last two weeks of pregnancy. Canad. Med. Assoc. J. 127, 377-380.
398
KEITH
BAILEY
D. A., MAHCOUB, A., SLOAN, T. P., IDLE, J. R., AND SMITH, R. L. (1980). A family and population study of the genetic polymorphism of debrisoquine oxidation in a white British population. J. Med. Genet. 17, 102-105. REINBERG, A., AND HALBERG, F. (1971). Circadian chronopharmacology Annu. Rev. Pharmacol. 11,455PRICE-EVANS,
492.
REINBERG, A., AND SMOLENSKY, M. H. (1982). Circadian changes of drug disposition in man. Clin. Pharmacokinet. 7, 40 I-420. RICHEY, D. P., AND BENDER, A. D. (1977). Pharmacokinetic consequences of aging. Annu. Rev. Pharmacol. Toxicol. 17, 49-65. RYLANCE, G. (1981). Drugs in children. &it. Med. J. 282, 50-51. TIMBRELL, J. A., MITCHELL, J. R., SNODGRASS,W. R., AND NELSON, S. D. (1980). Isoniazid hepatotoxicity: The relationship between covalent binding and metabolism in vivo. J. Pharmacol. Exp. Ther. 213, 364369. VESTAL, R. 358-382.
E. (1978). Drug use in the elderly: A review of problems and special considerations. Drugs 16,
YAFFE, S. J., AND JUCHAU, M. R. (1974). Perinatal pharmacology Annu. Rev. Pharmacol. 14,219-238.
TOXICOLOGY
AND
PHARMACOLOGY
Physiological
of Drug
(1983)
Factors Affecting KEITH
Bureau
3, 389-398
Research, Ottawa,
Health Ontario
Received
Drug Toxicity’
BAILEY Protection Branch, KlA OL2. Canada
TunneyS
Pasture,
Juiy 6. 1983
Physiological factors that affect the fate of drugs in the body and thereby have effects on their pharmacology and toxicology involve the systemsthat control absorption, distribution, metabolism, and excretion. The main factors are disease, genetics, and age. Nutritional status, sex, hormonal status (e.g., the effectsof pregnancy), and circadian rhythm have important influences. Maternal toxicity will affect the fetus. The absorption and excretion of drugs are frequently reduced by diseases. Excretion is reduced by impaired renal function, often found in the elderly, which may increase the toxic response. Distribution is affected by body weight and build, for example, the proportion of fat. The disposition of many drugs has been shown to be significantly influenced by circadian rhythms such that therapeutic doses may exhibit toxicity if administered at an inappropriate time of day. Metabolism is modified by environmental influences including previous food and drug experience, and various factors such as age, sex, and disease. Intersubject variations in drug disposition can be very great with possibly severe consequences for the individual; in this regard, knowledge of genetic polymorphism in drug metabolizing enzymes is rapidly increasing. The toxicology of a drug may be profoundly affected by a particular disease state, for example, the administration of a drug that might be a tumor promoter when a cancerous or precancerous condition exists. These effectsare illustrated with examples from the literature and recent studies undertaken in the Bureau of Drug Research.
Physiological factors that affect drug toxicity are multiple, interrelated in complex ways, and therefore difficult to unravel especially in human subjects. In this brief review of a very large topic it is intended to illustrate some of the effects that are known using examples from the literature and studies recently undertaken in the Bureau of Drug Research (BDR). AGE As yet, there are no studies that have followed a drug’s disposition in an individual from childhood into youth and adulthood, or from adulthood into old age. However, it is apparent that the clinical pharmacokinetics of many drugs are different in the newborn, the child, the adult, and the elderly. The physiological characteristics of ’ Adapted from a presentation at the February 21-23, 1983, meeting of the Toxicology Forum, Arlington, Va. 389
0273-2300183
$3.00
390
KEITH
BAILEY
TABLE
PHYSIOLOGICAL FACTORSCHARACTERISTIC
1
OF THE OLD
THAT AFFECT DRUG TOXICITY~
Absorption from GI tract possibly reduced Higher proportion of fat Reduced plasma protein binding Prolongation of plasma drug half-life (reduced metabolism?) Reduced renal elimination Multiple diseases Nutritional deficiencies Altered tissue response (receptor modification?) ’ Crooks (1976) Vestal (1978), Ricbey and Bender (1977).
aging listed in Table 1 influence drug disposition and have pharmacological consequences. Pathophysiological features of the very young (Table 2) and of infants and children and the consequent effects on drug efficacy and toxicity have been reviewed (Table 2). Morselli et al. (1980) presented detailed pharmacokinetic profiles of two dozen drugs that are commonly prescribed for children. An interesting relationship exists between theophylline and caffeine. Theophylline is used for the treatment of asthma in children and adults and for apnea and bradycardia in premature infants. In premature newborns it may be metabolized to caffeine which, like theophylline, is effective in recurrent apnea. Caffeine has an apparent plasma half-life of 43-231 hr compared with 12-64 hr for theophylline. It has been suggested (Morselli et al., 1980) that the concomitant presence of caffeine in premature infants could explain the lowered threshold for toxic effects observed with theophylline. The second example is the use of chloramphenicol in infants a few days old, where the half-life is about 26 hr in contrast to 4 hr in older children. This is now known to be due to the inability of neonates to glucuronidate the parent drug, but before this was realized, persistent high blood concentrations led to the deaths of infants who received a child’s dose. GENETICS The overall disposition of a drug depends on several processes that are under genetic control and possible toxic consequences are difficult to predict. Comparisons of the TABLE 2 PHYSIOLOGICALFACTORS CHARACTERISTIC OF THE VERY YOUNG THAT AFFECT DRUG TOXICITY a Gastric emptying prolonged Gastric pH neutral Higher proportion of body water Increased percutaneous absorption Reduced plasma protein binding Lowered hepatic microsomal activity Ready stimulation of metabolic capability (liver greater proportion of body mass) Lowered glucuronidation Reduced renal capability u Rylance (I 98 I), Morselli et at. (1980) Yaffe and Juchau (1974).
FACTORS
AFFECTING
391
DRUG TOXICITY
metabolism of drugs in populations that belong to different races or ethnic groups have revealed distributions of populations that handle drugs in distinctive ways. Table 3 shows the distribution of fast and slow acetylators of the antitubercular drug isoniazid (INH). Isoniazid has been extensively used among the autochthonic populations of North America alone in prophylaxis or in combination with other tuberculostats in treatment programs. Neuropathic toxic symptoms are more common with slow acetylators and signs of liver toxicity have been associated with fast acetylators; this point is somewhat controversial and it has been pointed out that homozygous and heterozygous rapid acetylators may differ in their susceptibilities to adverse effects (Kalow, 1982). Enzymes other than acetyl transferases are involved in INH toxicity. Thus, previous studies have attributed cell necrosis to the oxidative activation of the acetyl moiety of acetylhydrazine (Nelson et al., 1976; Timbre11 et al., 1980), and bisp-nitrophenyl phosphate is an amidase inhibitor that decreases the extent of hepatic necrosis induced by INH and acetyl-INH (Mitchell et al., 1976). In recent BDR studies, the specificity of hepatic microsomal amidase for the amide bonds found in INH and acetyl-INH was examined in male rats, guinea pigs, and rabbits of intermediate acetylator status. Rabbits were found to possess a metabolic profile reasonably similar to that of man. The rabbit showed the greatest hydrolytic activity for the amide bonds (Table 4). In each species the isonicotinoyl bond of acetyl-INH was the most susceptible. Furthermore, the rabbit amidase had a fourfold greater affinity for acetyl-INH than for INH (Michaelis constant data). Covalent binding studies with [‘4C]acetylhydrazine and [‘4C]acetyl-INH in rats and rabbits showed that there was less binding of the former, but more binding of [‘4C]acetylINH in the rabbit. Also, in the rabbit a greater proportion of the acetylhydrazine is metabolized to diacetylhydrazine. In rabbits, the most susceptible of the three species to INH hepatotoxicity, it seems to be the high amidase activity that produces higher amounts of acetylhydrazine and causes the high level of binding. Human amidase studies would therefore be very interesting. A well-known example of genetic control is the benzylic oxidation of &hisoquin investigated by Smith’s group in London (Price-Evans et al., 1980). Approximately 10% of the British population are homozygous for poor metabolizing capacity, 50% are homozygous rapid and 40% are heterozygous rapid metabolizers (Kalow, 1981). There was a 40-fold difference in individual dose requirements for a given therapeutic (antihypertensive) effect. TABLE 3 DISTRIBUTION(%)•
FACETYLATOR.SOFISONIAZID"
Fast Population
Slow
Heterozygous
Homozygous
Indian Caucasian Negro Eskimo Japanese Chinese
59 59 55 II 12 22
36 36 39 44 45 50
5 5 7 46 43 28
a See Kalow (1982).
KEITH
392
BAILEY
TABLE 4 HYDROLYTIC ACWITIES OFHEPATICMICROSOMAL PROTEIN'
Py *CO * NHNH-COCH, Py - CO-NHNH . COCH3 Py - CO-NI-INH2
Rat
Guinea pig
Rabbit
0.7 4.0 1.6
4.7 5.8 0.3
6.0 153 30
LIUnits are nmol/hr/mg protein (Thomas and Whitehouse, personal communication).
A recent report by Lennard et al. (1982) shows that the Pi-adrenoceptor antagonist metoprolol exhibits genetic polymorphism of the debrisoquin type. Plasma metoprolol concentrations were much higher in the poor hydroxylators, and this group had six times the area under the plasma concentration vs time curve (AUC) and three times the elimination half-life. Beta blockade was well maintained 24 hr after treatment in poor hydroxylators but was negligible in extensive hydroxylators. These workers suggested that asthmatic patients who are poor hydroxylators might be at some risk if treated with standard doses of metoprolol.2 Other associations with debrisoquin hydroxylation deficiency are shown in Table 5. It can be inferred that higher cancer rates are associated with a greater oxidative ability to produce active, toxic metabolites from various xenobiotics. There is scope for prospective studies along these lines. As part of BDR’s bioavailability assessment of quinidine formulations, McGilveray et al. ( 198 1) investigated pharmacokinetic variations in humans. Intersubject variation in the AUC was found to be high. The extremes of terminal half-lives among 24 subjects were 9.49 and 6.24 hr, the plasma quinidine concentrations at 24 hr differing by a factor of 10. In the BDR study, the frequency distribution of clearance appeared to be normal. The plasma binding of disopyramide (DP) and mono-Wdealklyldisopyramide (MND) have been reported recently by Bredesen et al. (1982). The drug and its metabolite have antiarrhythmic effects. Plasma from 70 patients on maintenance therapy but otherwise drug free was analyzed for the two substances. A wide variation was found: from 1 to 20 ctmol/liter for DP and from 1 to 16 ~mol/liter for MND and MND:DP ratios of 1.6-4.0. Partial data from the plasma of healthy subjects are shown in Table 6. It was found that the unbound fraction of both DP and MND increased with higher concentrations and, further, that the unbound fraction of each one was increased in the presence of clinically important concentrations of the other. According to these authors, the interpatient variance of unbound DP concentration might be tenfold or higher in the assumed therapeutic range. Unbound drug and metabolite concentrations should be monitored in order to minimize toxic effects and maximize the therapeutic benefit. The use of drugs with such characteristics in patients with cardiac disease is complicated by the affect of the disease itself on disposition. * This matter has received further attention very recently; see D. B. Jack, M. J. Kendall, M. Wilkins, and C. P. Quarterman (1983). N. Engl. J. Med. 308,964; P. Dayer, F. Courvoisier, L. Balant, and J. Fabre (1983). N. Engl. J. Med. 3B8, 964-965; M. S. Lennard, S. Freestone, L. E. Ramsay, G. T. Tucker, H. F. Woods, and J. H. Silas (1983). N. Engl. J. Med. 308, 965.
FACTORS
AFFECTING
393
DRUG TOXICITY
TABLE 5 DECREASED
DEBRWQUINE
METABOLISM
Decreased metabolism associations Sparteine Guanoxan Phenacetin Perhexiline Nortriptyline Phenformin Propranolol Metoprolol Phenytoin Mephenytoin
AWXIATIONS”
Normal metabolism associations Amobarbital Antipyrine Tolbutamide Clozapine Possible decrease associations Liver cancer (? aflatoxin) Bronchial carcinoma
a cf. Kalow (198 I).
DIETARY
CONSIDERATIONS
Dietary habits affect body weight (proportion of fat, enzyme levels, etc.), gut pH, flora, transit times, and so forth, and so considerably modify pharmacological responses (Fig. 1). Thus, investigations into the hepatotoxicity of CC& indicated that greater toxicity arose among animals receiving a lower proportion of carbohydrate but increased fat and protein in isocaloric diets (Nakajima et al., 1982). On a superficial level, if a drug is taken with meals the outcome can be very different from that following a period of starvation. For example, propranolol qid followed by meals produced maximum plasma levels and AUCs that were twice those in fasting subjects. It was suggested that food decreases presystemic elimination by increasing the rate of blood flow in the viscera. (Dvomik et al., 1982). During investigations into the bioavailability of metronidazole formulations, McGilveray et al. (1978) noted that a reference solution gave a lower bioavailability in starved subjects than did a reference tablet. It was suggested that in some individuals, gastrointestinal flow was so rapid TABLE 6 DISOPYRAMIDE
(DP)
AND
MONODEALKYLDISOPYRAMIDE FROM FIVE HEALTHY SUBJECTS”
(MND)
Mean equilibrium (mol/U DP initial (wmol/l) 1
5 10 15 20 ’ The assumed therapeutic
0.8 4.0 7.0 10.6 12.8 3.8 12.4 range
Mean
0 0 0 0 0 14.4 14.3 of DP
is 6-15
PLASMA
DP unbound fraction (‘%)
MND
DP
IN SPIKED
rmol/l
+ SD
12* 19f 28k 34+ 43i 35 Ik 54k (Bredesen
et al.,
2 5 4 5 8 10 10 1982).
Range 10-14 14-28 22-3 1 29-41 34-50 25-46 40-64
394
KEITH
BAILEY
FOOD
GENERAL
DRUG-PROTEIN
NUTRUTION
PH
ACCUMULATION OF LIPOPBILIC DRUGS I ACIDIC/BASIC DRUG ABSORPTION, ELIMINATION
BINDING I ENZYME LEVELS, BIOTRANSFORMATION, ELIMINATION
MOBILIZATION
MOBILIZATION -.l\/--MODIFIED
EFFECTS
FIG.1. Physiological associations of food intake with drug pharmacology and toxicology.
that solutions rapidly passed the absorption sites. The disintegrating tablets moved more slowly in the area of absorption, with particles being held in the mucosal layer. Feldman et al. (1982) showed that theophylline kinetics in children are affected by the diet. The half-life was 6.8 hr on a normal diet, 4.8 hr on a high protein diet, and 18.1 hr on a high carbohydrate diet. Significant changes in serum levels could occur in l-2 weeks. PHYSIQUE From diet we make a natural progression to physique. A curious example was provided recently of an apparent sex difference in dapsone treatment of leprosy in Ethiopian men and women (Modderman et al., 1982). These patients received 900mg injections into the buttocks using a 3.5-cm-long needle. Serum concentrations reached 2 pg/ml on Day 3 in men, declining to about 0.5 pg by Day 10 and declining steadily to 0.1 pg/ml by Day 28. In women, the concentration was 1 &ml on Day 3, declining steadily to 0.4 pg/ml by Day 28. Almost certainly, the women received the injections into the subcutaneous fat and the men received injections into the muscle. Cockshott and Thompson (1982) pointed out that the sexual dimorphism in total fat content (women 24%, men 17%) extends to its distribution. Pregnancy is a condition that as yet remains sex specific! One of its effects is to increase glomerular filtration rate by 30-50% and renal plasma flow by about 25%, values returning to normal soon after delivery (Dvorchik, 1982). Therefore, dosages of drugs that are cleared by renal mechanisms may need to be adjusted during and after pregnancy if toxic effects are to be avoided. An instance of lithium toxicity due to failure to make a dose adjustment postpartum has been attributed to these changes.
FACTORS
AFFECTING
395
DRUG TOXICITY
Parsons and Pelletier (1982) suggested that inhibition of the hepatic mixed-function oxidase system could be responsible for the delayed elimination of caffeine that was observed near term, where levels may be three times normal. The effects of posture on pharmacokinetic parameters of drugs have been studied by several groups. This simple physiological variable can also change the toxicological profile. A recent report by Heller et al. (1982) associated the prior consumption of nonsteroidal anti-inflammatory drugs (NSAIs) with the development of benign esophageal stricture, a consequence of simple mechanics. NSAIs have long been associated with gastric erosions. Hey et al. (1982) investigated the esophageal transit of common types of tablets and capsules. They subsequently recommended that subjects should swallow the preparation with at least 100 ml of water and remain standing for at least 90 sec. CIRCADIAN
RHYTHMS
It seems to be popular for people to blame a maliferous phasing of their biorhythms when everything seems to go wrong at once. Circadian pharmacokinetics have been demonstrated for the drugs and related xenobiotics shown in Table 7. Cisplatin provides a particularly interesting example. This very useful but unfortunately poorly tolerated anticancer drug shows the lowest nephrotoxicity when given in the evening, which is when the urinary concentration is lowest. The highest antitumor activity is simultaneous with best tolerance. Circadian rhythms provide yet another complicating factor in evaluating toxic responses. DISEASE Although disease is a pathological state it is appropriate to consider it in the context of physiological factors. There are numerous toxicological effects to be expected when drugs are administered to a diseased patient in comparison with results from healthy subjects in whom drugs are tested. In general, the metabolic clearance may be reduced if liver function is impaired and thus total plasma clearance reduced even when the renal clearance is normal. Drug responses are increased when there is renal failure since excretion is reduced. The plasma levels of drugs with high liver extraction ratios may be increased when patients are hypotensive. TABLE 7 XENOBIOTICS
Acetaminophen Aspirin Digitalis Hexobarbital Nortriptyline Potassium chloride Sulfasymazine Theophylline
EXHIBITING
CHRONOPHARMACOKINETICS’
Aminopyrine Cisplatin Erythromycin Indomethacin Phenacetin Sodium salicylate Sulfonilamide Xylose
Ampicillin Clorazepate Ferrous sulfate Lithium carbonate Phenytoin Soman Sulindac
a Reinberg and Smolensky (1982), Elsmore (198 I), Reinberg and Halberg (197 1).
396
KEITH
BAILEY
TABLE 8 IN VITRO GENOTOXKITY TESTINGOFDIAZEPAM~ 1. Diazepam did not induce DNA repair in freshly isolated rat hepatocytes in subtoxic dose range. 2. Diazepam did not alfect DNA repair induced by methylmethane sulfonate in freshly isolated rat hepatocytes. 3. Mutations by 8-azaguanine at HGPRT locus in rat liver cell line T5lB were not induced by diazepam, with or without metabolic activation. 4. Diazepam did not induce DNA synthesis in quiescent cells (low Ca). b 5. Diazepam and oxazepam, with or without metabolic activation, were unable to induce reverse mutations in Salmonella typhimurium strains TA 98 and TA 100. 6. As in 5 (above) except with a histidine-enriched
medium (Batzinger et al., 1978).
’ Matula and Swierenga (personal communication). b Boynton and Whitfield (1980).
Studies recently carried out in our laboratories on diazepam represent a special example of the effect of disease on possible drug toxicity. Diazepam is one of the most commonly prescribed tranquilizers, and a great deal of controversy arose following some suggestions (Horrobin et al., 1979; Hoi-robin, 198 1) that it might be a promoter of breast cancer in women. No evidence has accrued from epidemiological studies that diazepam is a tumor promoter. We undertook in vitro studies designed to determine genotoxic activity, with the results shown in Table 8. (The stimulation of DNA synthesis in calcium-deprived cells is a short-term promoter assay described by Boynton and Whitfield (1980) Canadian National Research Council.) In the meantime, in vivo studies in female rats induced for mammary tumors by the administration of dimethylbenzanthracene were started according to the protocol shown in Table 9. This pilot study provided no evidence for the promotion of palpable tumors. This last example is a rather particular one of drug toxicity that could be sex and disease related (and because of the type of disease, also age related). TABLE 9 SUMMARYOFDIAZEPAM PILOTSTUDYINFEMALERATS" Virgin female Sprague-Dawley rats purchased at 25 days 54 days old: 5 mg dimethylbcnzanthracene in 1.0 ml corn oil, gavage Group 1 (control, 19): vehicle (gum tragacanth) Group 2 (9): diazepam 1.0 mg/kg/day, SX/week Group 3 (8): diazepam 10.0 mg/kg/day, SX/week Mammary regions palpated weekly All surviving animals killed after 45 weeks and autopsied Employ life table analysis to diminution of prolongation of time to tumor. u Matula (personal communication).
FACTORS
AFFECTING
DRUG
397
TOXICITY
ACKNOWLEDGMENTS The comments and advice of G. L. Frederick, T. Matula, Thomas, and L. W. Whitehouse are greatly appreciated.
I. J. McGilveray,
S. H. H. Swierenga,
B. H.
REFERENCES BATZINGER, R. P., BUEDING, E., REDDY, B. S., AND WEISBURG, J. H. (1978). Formation of a mutagenic drug metabolite by intestinal microorganisms. Cancer Res. 38, 608-6 12. BOYNTON, A. L., AND WHITRELD, J. F. (1980). Stimulation of DNA synthesis in calcium-deprived T5 1B liver cells by the tumour promoters phenobarbital, saccharin, and 12-O-tetradecanoylphorbol13-acetate. Cancer Res. 40,4541-4545. BREDESEN, J. E., PIKE, E., AND LUNDE, P. K. M. (1982). Plasma binding of disopyramide and mono-Ndealkyldisopyramide. Brit. J. Clin. Pharmacol. 14, 673-676. COCKSHOTT, W. P., AND THOMPSON, G. T. (1982). Injections into fat instead of muscle. N. Engl. J. Med. 307, 1581. CROOKS, J., O’MALLEY, K., AND STEVENSON, I. H. (1976). Pharmacokinetics in the elderly. Clin. Pharmacokinet. 1, 280-296. DVORCHIK, B. H. (1982). Drug disposition during pregnancy. Zntl. J. Biol. 3, 129-137. DVORNIK, D., KRAML, M., DUBUC, J., FENCIK, M., WEIDLER, D., AND MULLANE, J. F. (1982). Propranolol concentrations in healthy men given 80 mg daily in divided doses: Effect of food and circadian variation. Curr. Ther. Res. 32, 214-224. ELSMORE, T. F. (198 1). Circadian susceptibility to soman poisoning. Fundam. Appl. Toxicol. 1.238-24 1. FELDMAN, C. H., HUTCHINSON, V. E., SHER, T. H., FELDMAN, B. R., AND DAVIS, W. J. (1982). Interaction between nutrition and theophylline metabolism in children. Ther. Drug Monit. 4,69-76. HELLER, S. R., FELLOWS, 1. W., OGILVIE, A. L., ANLI ATKINSON, M. (1982). Non-steroidal anti-inflammatory drugs and benign oesophageal stricture. Brit. Med. J. 285, 167-168. HEY, H., JORGENSEN, F., WRENSEN, K., HASSELBACH, H., AND WAMBERG, T. (1982). Oesophageal transit of six commonly used tablets and capsules. Brit. Med. J. 285, 17 17- 17 19. HORROBIN, D. F., GHAYUR, T., AND KARMALI, R. A. (1979). Mind and cancer. Luncet 1, 978. HORROBIN, D. F. (1981). Diazepam as tumour promoter. Lancef 1, 277-278. KALOW, W. (198 1). The metabolism of xenobiotics in different populations. Canad. J. Physiol. Pharmacol. 60, l-12. KALOW, W. (1982). Ethnic differences in drug metabolism. Clin. Pharmacokinet. 7, 373-400. LENNARD, M. S., SILAS, J. H., FREESTONE, S., RAMSAY, L. E., TUCKER, G. T., AND WOODS, H. F. (1982). Oxidation phenotype-A major determinant of metoprolol metabolism and response. N. Engl. J. Med. 307, 1558-1560. MCGILVERAY, 1. J., MIDHA, K. K., Loo, J. C. K., AND COOPER, J. K. (1978). Bioavailability of commercial metronidazole formulations. Intl. J. Clin. Pharmacol. 16, 110-I 15. MCGILVERAY, I. J., MIDHA, K. K., ROWE, M., BEAUDOIN, N., AND CHARETTE, C. (1981). Bioavailability of 11 quinidine formulations and pharmacokinetic variations in Humans. J. Pharm. Sci. 70, 524-529. MITCHELL, J. R., ZIMMERMAN, H. J., ISHAK, K. G., THORGEIRSSON, U. P., TIMBRELL, J. A., SNODGRASS, W. R., AND NELSON, S. D. (1976). Isoniazid liver injury: Clinical spectrum, pathology, and probable pathogenesis. Ann. Intern. Med. 84, 18 l-192. MODDERMAN, E. S. M., HILBERS, H. W., ZUIDEMA, J., MERKUS, F. W. H. M., AND WARNDORFF, T. (1982). Injections into fat instead of muscle. N. Engl. J. Med. 307, 1581. MORSELLI, P. L., FRANC~MORSELLI, R., AND Bossr, L. (1980). Clinical pharmacokinetics in newborns and infants. Clin. Pharmacokinet. 5, 485-527. NAKAJIMA, T., KOYAMA, Y.. AND SATO, A. (1982). Dietary modification of metabolism and toxicity of chemical substances-With special reference to carbohydrate. B&hem. Pharmacol. 31, 1005-1011. NELSON, S. D., MITCHELL, J. R., TIMBRELL, J. A., SNODGRASS, W. R., AND CORCORAN, G. B. (1976). Isoniazid and iproniazid: Activation of metabolites to toxic intermediates in man and rat. Science 193, 901-903. PARSONS, W. D., AND PELLETIER, J. G. (1982). Delayed elimination of caffeine by women in the last two weeks of pregnancy. Canad. Med. Assoc. J. 127, 377-380.
398
KEITH
BAILEY
D. A., MAHCOUB, A., SLOAN, T. P., IDLE, J. R., AND SMITH, R. L. (1980). A family and population study of the genetic polymorphism of debrisoquine oxidation in a white British population. J. Med. Genet. 17, 102-105. REINBERG, A., AND HALBERG, F. (1971). Circadian chronopharmacology Annu. Rev. Pharmacol. 11,455PRICE-EVANS,
492.
REINBERG, A., AND SMOLENSKY, M. H. (1982). Circadian changes of drug disposition in man. Clin. Pharmacokinet. 7, 40 I-420. RICHEY, D. P., AND BENDER, A. D. (1977). Pharmacokinetic consequences of aging. Annu. Rev. Pharmacol. Toxicol. 17, 49-65. RYLANCE, G. (1981). Drugs in children. &it. Med. J. 282, 50-51. TIMBRELL, J. A., MITCHELL, J. R., SNODGRASS,W. R., AND NELSON, S. D. (1980). Isoniazid hepatotoxicity: The relationship between covalent binding and metabolism in vivo. J. Pharmacol. Exp. Ther. 213, 364369. VESTAL, R. 358-382.
E. (1978). Drug use in the elderly: A review of problems and special considerations. Drugs 16,
YAFFE, S. J., AND JUCHAU, M. R. (1974). Perinatal pharmacology Annu. Rev. Pharmacol. 14,219-238.