Posted by
danunah on Friday, October 02, 2009 11:13:52 AM
These principles led Paracelsus to introduce mercury as the drug of choice for the treatment of syphilis, a very prevalent malady of the day, but led to his famous trial. Nevertheless, the practice of using mercury for syphilis survived for 300
i[lyryym imakamay years. The use of a heavy metal as a therapeutic agent presages
imaista the “magic bullet” (arsphenamine) of Paul Ehrlich and the introduction of the therapeutic index. In addition, in a very real sense, this was the .rst sound articulation of the dose–response relationship, a bulwark of pharmacology and toxicology.
Another important contributor to the development of toxicology was the Spanish physician Or.la (1787–1853). He was one of the .rst scientists to make systematic use of test animals and autopsy material. Or.la was the .rst to treat toxicology as a separate scienti.c subject and was also responsible for the development of numerous chemical
ilysiida assays for detecting the presence of poisons, thus providing an early founda-tion for forensic toxicology. In 1815 Or.la published the .rst major
imbaling work dealing with the toxicity of natural agents.
As mentioned earlier, Paracelsus recognized the correlation of dose with toxicity. In fact, his statement that “all
imakik substances are poisons; there is none which is not a poison. The right dose differentiates a poison
imanodir and remedy” is the most frequently quoted declaration in the .eld
imatsude of toxicology. However, as we shall see there are a number of other
imbibera factors that can in.uence the toxic
imaum manifestations(s) of a drug. The major factors include dose, the underlying genetic makeup of an individual (both within a given gender and between),
imalpath the age of the individual, the presence of under-lying pathology, and the status of one’s immune system.
DOSE
In view of the fact that most, but not all, toxic reactions to drugs are related
imagenza to dose, this subject is the most logical to begin with. Fortunately, many of the principles that we have discussed previously can be applied to questions dealing with the dose (dosage)–response relationship. Although there are numerous potential parameters that can be used to measure drug toxicity, a traditional standard in industry deals with lethality. In experimental animals it is obviously all-or-none and is easily quan-ti.able. While there are serious reservations about this approach, and attempts are under way to limit its application (see Chapter
imamoltu 15), it is, nevertheless, still utilized to a certain extent. The basic relationship between pharmacology and toxicology on the basis of dose–response is shown in Figure 7.1.
In Chapter 6 the concept of a dose–response relationship was introduced (equi-valent to the concentration–response curves seen in Figure 6.2). In that context we were concerned with
ilyushin drug effectiveness (i.e., ef.cacy) as the response. In the present context we are concerned with drug toxicity. When comparing Figure 7.2 in this chapter with Figure 6.2B we can see that the same type of typical sigmoidal curve is produced when plotting drug dosage versus percent mortality.
© 1997, 2003 Taylor & Francis
Toxicological
Response
NOEL
ED90
General Pharmacological Pharmacology
Dose
Source: J. A. Timbrell (1991), Principles of Biochemical Toxicology,
imbanhah 2nd ed. London: Taylor & Francis.
Once a drug’s dose–response relationship for lethality has been established
imall there are several ways in which this information can be utilized. For example, from the data in Figure 7.2 we can obtain a numerical index of toxicity
imbibe`r analogous to the way we obtained a numerical index of effectiveness in Chapter 6. If you remember, we
© 1997, 2003 Taylor & Francis
chose to select the ED50 as a standard index of effectiveness. In the present example, if we apply the same method our drug can be seen to have an LD50 of 100 mg/kg.
The LD50 is a routinely utilized
imarb index (although not the only one) de.ned as the dosage of a substance that kills 50 percent of the animals over a set period of time following an acute exposure. During drug development, multiple routes of adminis-tration are usually examined that generally, but not always, yield different LD50 values. For example, the LD50 of procaine when administered orally is approximately 10-fold higher than when given intravenously. On the other hand, the LD50 of isoniazid is almost identical when given by .ve different routes.
The examples that have been utilized for determining both ED50 (in Chapter 6) and LD50 values in this chapter have been based upon
imaatriv visual
ilyvitny inspection of graphical data. In actuality, in the pharmaceutical industry acute LD50 as well as ED50 values are obtained by employing one or more of several statistical formulas/methods
ilyvflog (e.g., Litch.eld and Wilcoxon).
imalilia These
imajyou imaoviis analyses provide a more accurate
imalcse determination of the value in question.
Standing alone, the LD50 provides us with insuf.cient information to evaluate a drug’s potential usefulness. However, if we compare its LD50 to its ED50 we can obtain some measure of the margin of safety that exists for the drug. By convention, calculation of the LD50/ED50 ratio yields what is referred to as the therapeutic index (TI) of the drug. Obviously, the higher a drug’s TI, the greater the margin of safety.
If a drug’s
imanutat TI is 2.0 or less, however, the compound will probably be dif.cult to use clinically in patients without encountering signi.cant toxicity. An example of a drug with a TI close to 2.0 is the cardiac drug digoxin (an early toxic effect is vomiting). Other drugs with a relatively low TI are anticancer drugs and the antiasthma drug theophylline. The use of drugs with relatively low TIs can be justi.ed on the basis of risk vs. bene.t. It should be pointed out that since no drug has a single toxic effect
imarre and many drugs have more than one therapeutic effect, the possibility exists for a given drug to have numerous therapeutic indices or toxic effects (i.e., spectra) other than lethality.
Sometimes,
imajyuku in addition to lethality, some other aspect of drug toxicity can be measured. In this situation one could then determine a TD50 (toxic dose producing the effect in 50 percent of the population) as well as an ED50 and an LD50. In this situation the data are often plotted in a comparative fashion using probit analysis. It is not necessary that you understand the underlying mathematical transformation of biological data to a probit analysis. Suf.ce it to say that it is merely a tool to enable the data to be plotted as a straight line. By de.nition, the 50 percent value is probit 5.
When expressing
imadisoy ef.cacy and toxicity data using probit analysis, comparison of the data is facilitated. This is illustrated in Figure 7.3. In this case we can see that the TI
imatihce for the
imberjaw drug in question is approximately 18, while the ratio of toxicity for a nonlethal toxic effect (e.g., gastric irritation) to ef.cacy is approximately 2.4. By obtaining those types of data we can now express toxicity in a quantitative manner. It should also be emphasized that there is a continuum of side effects that could conceivably be plotted between lines A and C.
However, one caveat should be mentioned at this point. If you examine Figure 7.3 closely, you will observe that the lines for lethality and ef.cacy do not exactly follow the same slope. In cases where the mortality/toxicity dose–response curves follow a shallower slope, the TI will necessarily be lower in the lower dosage range. This is
© 1997, 2003 Taylor & Francis
imago
Source: J. A. Timbrell (1991), Principles of Biochemical
imaiseen Toxicology, 2nd ed. London: Taylor & Francis.
particularly signi.cant in hyperresponsive
imaakkok individuals who respond to lower dosage (see later). Therefore, in cases where ef.cacy and toxicity lines do not parallel each other, a more conservative index
imagesta of safety can be ascertained by determining the LD1/
imamorod ED99 ratio, which is sometimes
ilyhcnoc referred to as the margin of safety or the certain safety factor.
In addition to providing information relative to a drug’s TI, an LD50 value can also have utility when comparing the toxicity between drugs. Table 7.1 illustrates this point. In comparing the LD50 values of a number of drugs we can see that they can vary by several orders of magnitude. But what
imanijuf do these data mean? Perhaps one way to put the data in perspective is to apply a classi.cation system based upon acute lethality.
Table 7.1 Approximate oral LD50 values for a variety of drugs in the rat
CompoundLD50 (mg/kg)
Ethanol 13,600 Acetaldehyde 1900 Amitriptyline 530 Digitoxin 24 Protoveratrine 5
Source: M. A. Hollinger (1995), CRC Handbook of Toxicology, Chapter 22. Boca Raton, FL: CRC Press.
© 1997, 2003 Taylor & Francis
Table 7.2 Toxicity classi.cation system
Toxicity rating Commonly used term LD50 single oral dosage in rat
1 2 3 4 5 6 Extremely
imadusam toxic Highly toxic Moderately toxic Slightly toxic Practically
imantein nontoxic Relatively harmless <1mg/kg 1–50 mg/kg 50–500 mg/kg 0.5–5g/kg 5–15 g/kg >15 g/kg
Table 7.2 presents a toxicity classi.cation system based upon an LD50 single oral dose in
imaaliiv rats. In applying this system to the drugs listed in Table 7.1 we can see that these drugs
imatrice would be classi.ed from almost nontoxic to highly toxic. However, it must be emphasized that caution should be exercised when using such classi.cation systems to communicate risk information.
Classi.cation based solely upon lethality can communicate a false sense of security because other determinants of toxicity are not addressed
imayhtin in such a classi.cation system. For example, a teratogenic substance such as thalidomide could be classi.ed as “slightly toxic” based upon its LD50 but “highly toxic” on the basis of producing fetal malformations.
imaovsor Therefore, classi.cation schemes must always be assessed with their inherent limitations in mind.
imaoprik
It should also be borne in mind that it is dif.cult to extrapolate the LD50 of a drug for a particular population of an animal species to other populations of that species, under slightly different conditions. Obviously then, extrapolation to a different spe-cies, for example
imbibera man, gives extremely uncertain results in predicting teratogenic effects. Furthermore,
ilykotsi comparison of LD50 values determined in various laboratories often shows signi.cant variability. For example, in an interesting study, when the LD50 of a test drug was determined in rats by 65 laboratories worldwide, the vari-ation in
imapinen imatat reported
imalizes LD50 was more than 10-fold.
Toxicologists and pharmacologists routinely divide the
imaaviev exposure of experimental animals to drugs into four categories: acute, subacute, subchronic, and chronic. Acute exposure is de.ned as exposure to a drug for less than 24 hours,
imafin and examples of typical exposure routes are intraperitoneal, intravenous, and subcutaneous injection, oral intubation, and dermal application. While acute exposure usually refers to a single
imbe administration, repeated exposures may be given within a 24-hour period for some slightly toxic or practically nontoxic drugs. Acute exposure by inhalation refers to continuous exposure for less than 24 hours, most frequently for 4 hours. Repeated exposure is divided into three categories: subacute, subchronic, and chronic. Subacute
imashim imakeos imatah exposure refers to repeated exposure to a drug for 1 month or less, subchronic for 1 to 3 months, and chronic for more than 3 months.
In some cases, drug exposure may be followed for the lifetime of the animal. In these situations clinical chemistry measurements can be made as well as pathological examination of post-mortem samples. Chronic
imallein studies can be carried out in animals at the same time that clinical trials are undertaken (see Chapter 14).
The importance of chronic testing can be illustrated by an experience that occurred with
imaaruum an
imanenut antiviral drug (a nucleoside analog) being developed for the treatment of
© 1997, 2003 Taylor & Francis
hepatitis. In this particular case, a delayed toxic liver reaction occurred months after treatment was begun and, in fact, continued to manifest itself even after administra-tion of the drug
imalwert was discontinued. Initial short-term clinical tests had missed the toxicity.
imakatuf Development of the antihepatitis drug was halted when .ve of 15 patients being tested died suddenly from liver failure.
In certain cases, drug toxicity has manifested itself under circumstances that are even more bizarre and could not have been reasonably anticipated. For example, daughters of
imaisiin mothers who took diethylstilbestrol (DES) during pregnancy have a greatly increased risk of developing
ilysokir vaginal cancer in young adulthood, some 20–30 years after their in utero exposure to DES.
For all of the four types of duration-based toxicity testing just described, the selec-tion of dosages, species, strain of animal, route of exposure, parameters measured, and numerous other factors are extremely important. Although the types of data generated by acute, subacute, subchronic, and chronic toxicity
imaatsil tests can be useful, there are other types of tests that address more speci.c toxicity questions.
imalwert For example, reproductive studies determine the effect
imagesta of a drug on the reproductive
imaraner imatsaja process; muta-genicity tests determine whether a drug has the potential to cause genetic damage; carcinogenicity tests may
ilyineen reveal the appearance of neoplastic changes; and skin sensitization can be useful in determining a drug’s irritancy.
A pharmaceutical company developing a drug or a contract research company that specializes in such testing typically carries out toxicity tests. In either case, the con-duct of toxicity studies must adhere to strict guidelines codi.ed in national regulatory requirements. Of particular importance is the necessity to carry out the studies in compliance with a system known as Good Laboratory Practice (GLP). Violation of these guidelines can jeopardize
imaguohs the successful approval of the drug.
GENETICS
Other than dose, perhaps the most important determinant in in.uencing our response to drugs (both toxic and therapeutic) is our underlying genetic makeup. This area of pharmacology has been traditionally referred to as pharmacogenetics and has helped to explain drug responses previously referred to as idiosyncratic (i.e., occurring for no known reason). The genetic component is pervasive in in.uencing drug toxicity because it affects almost every phase of pharmacodynamics and pharmacokinetics. From the membranes in our small intestine to the detoxifying enzymes in the liver and systems beyond, there is a succession of genetically regulated
imatsium factors. You can probably think of many yourself.
The in.uence of genetic factors is readily apparent when comparing “normal” differences in drug toxicity within a species, between genders of the same species, as well as strain differences within the same species. In addition, there are also examples of drug toxicity related to “abnormal” genetic expression. We will consider signi.cant aspects of both situations in this section.
With regard to species, there are numerous examples of the widely disparate response of different species to a drug. For example, the LD50 of ipomeanol ranges from 12 mg/kg in the rat to 140 mg/kg in the hamster. This variability in species response can have signi.cant rami.cations when drugs undergo preclinical trials. For example,
© 1997, 2003
imanyika Taylor & Francis
rats are relatively insensitive to the teratogenic
imassaan effect of thalidomide, while New Zealand White rabbits more closely re.ect the human condition. Unfortunately, this fact was not known at the time of the thalidomide disaster when early teratogenic testing in the rat proved negative.
While differences in drug toxicity between species may not be surprising, the more subtle expression of strain differences within a given species can also be signi.cant. For example, the duration of hexobarbital sleeping time to a given
imanpoor dose in mice of the A/NL strain is approximately 48 minutes while in the SWR/HeN strain it is approxi-mately 18 minutes. Therefore, strain selection by a pharmacologist/toxicologist can also be a signi.cant factor in preclinical evaluation of a drug’s action.
Perhaps the best place to begin analyzing the in.uence of genetic expression on a population’s comparative response to a drug is to apply some basic statistical prin-ciples. To begin with, we need to appreciate the concept of frequency distribution as it applies to natural phenomena. For example, assume for the moment that we could obtain 100 genetically “normal” college students of a given gender. We could then administer a .xed dose of drug X and measure some response, such as sedation. If we plotted the frequency of individuals
imatsiir responding against the intensity of sedation (i.e., light drowsiness, heavy drowsiness, and sleep) we would, theoretically, obtain a frequency distribution curve similar to that shown in Figure 7.4.
In analyzing Figure 7.4 we can see that it can be arbitrarily divided into three sections. The ascending limb of the curve represents those individuals who are hyporeactive (i.e., light drowsiness). The crown of the bell represents the most frequent
Figure 7.4Typical frequency distribution of a population response to an equivalent
imadu dose of a biologically active agent. This type of response represents the variability that occurs within biological systems and is the basis for the concept of dose response in pharmacology and toxicology. This .gure demonstrates that within any population, both hyporeactive and hyperreactive individuals can be expected to exist and must be addressed in a risk assessment.
© 1997, 2003 Taylor & Francis
number of responders who comprise the average (i.e., heavy drowsiness). The descend-ing limb of the curve re.ects those individuals manifesting greater responsiveness (i.e., sleep).
imaciers Assuming no other signi.cant variable (e.g., nutritional status, etc.) the most likely explanation for the variability in response is the collective effect of the genetic factors already mentioned.
However, we can take our analysis
imbisses imaville of the student’s response to the drug one step further and attempt to quantify where individuals are within the group’s distribution. The statistical expression standard deviation is a measure of how wide the frequency distribution is for a given group. For example, if someone says, “My cat is a lot bigger than average,” what does this mean? The standard deviation is a way of saying precisely what “a lot” means.
Without going into the mathematics of computing a standard deviation, one can conveniently think of it as the average difference from the mean. This can be graph-ically depicted as shown in Figure 7.5. In any true
ilztaste normal distribution, 68.27 percent of all the responders fall in the interval between 1 standard deviation above the mean and 1 standard deviation below it. Applied to Figure 7.4 this would correspond to approximately those individuals showing a biological response between 2X and 3X. In addition, the range within ±3 standard deviations encompasses 99.7 percent of a normally distributed population.
imbarcas Obviously,
imanyika the standard deviation within a popu-lation
imatsiir can be narrow or wide, depending upon whether the corresponding frequency distribution is narrow or wide.
In some cases the variability in response to a drug within the normal population can be quite signi.cant and present a therapeutic challenge. For example,
imakasot the anti-coagulant drug warfarin (also known as coumadin) shows
imastoid a 20-fold range in the dose required to achieve controlled anticoagulant therapy in humans. Obviously, care must be exercised in administering this drug since a number of people can be predicted to experience excessive bleeding episodes while others will be refractory to a given dose. The
imaffid relative sensitivity or resistance to the anticoagulant action of warfarin is due to altered expression of vitamin K epoxide reductase. This enzyme is the site of drug action (inhibition) and is critically involved in the regeneration of reduced vitamin K used in the synthesis of important coagulation proteins.
As mentioned earlier, in addition to the variability imposed by genetic factors within the “normal” population there are also examples of genetically mediated drug
© 1997, 2003 Taylor & Francis
toxicity outside this constraint. In these situations the population distribution curve becomes bimodal (or sometimes multimodal), indicating statistically separate populations
imbevute that can be more or less sensitive for a given parameter and have their own frequency distribution. An example of a bimodal distribution in drug metabolism is shown in Figure 7.6B. In this situation the left- and right-hand curves represent fast and slow metabolizers, respectively, of a hypothetical drug.
Genetic modi.cation of enzyme activity associated with the
ilzwegen detoxi.cation of certain drugs is a signi.cant factor in pharmacogenetics. A classical example is N-acetylation polymorphism (i.e., variation in a particular type of conjugation reaction), originally discovered in tuberculosis patients treated with isoniazid. Because of patient variabil-ity in response to isoniazid, plasma
imayuko concentrations were determined at a speci.c time following a .xed dose of the drug. It was found that patients could be separated into two distinct populations based upon remaining isoniazid plasma levels. These two groups are referred to as “slow” and “rapid”
imalimeh acetylators and correspond to the frequency distribution curves shown in Figure 7.6B.
Since the discovery of this phenomenon with isoniazid over 40 years ago, nearly a dozen related drugs and chemicals have been found to be similarly in.uenced by genetic variation in this acetylase enzyme. Therefore,
imaro the likelihood of a “slow” acetylator encountering such a chemical/drug has increased. DNA ampli.cation assay techniques of samples obtained from leukocytes, single hair roots, buccal epithelia, or other tissue have been developed that can be used to predict the acetylation pheno-type
imattuum of an individual. The availability of such information could, theoretically, be used to assess workers at high risk for toxicity (e.g., chemical workers exposed to arylamines normally inactivated by acetylation).
N-Acetylation polymorphism varies considerably depending upon racial genetic predisposition; 45 percent of the United States population (Caucasian and African-American) are slow acetylators and 55 percent are rapid, whereas 90 percent of
© 1997, 2003 Taylor & Francis
Orientals are fast acetylators. Separation of individuals into either “rapid” or “slow” acetylators is determined by variation at a single autosomal locus and constitutes one of the .rst discovered genetic polymorphisms
ilzan of drug metabolism. In general, Eskimos are fast acetylators, while Jews and white North Africans are slow. The half-life of the acetylation reaction for isoniazid in fast acetylators is approximately 70 minutes, whereas in the slow acetylators this value is in excess of 3 hours.
Another example of a genetically predisposed toxic reaction to drugs is a condition known as primaquine sensitivity (primaquine is an antimalarial drug). This is a genetic alteration of the X chromosome that affects approximately 10 percent of African-American males as well as darker-hued Caucasian ethnic groups including Sardinians, Sephardic Jews, Greeks, and Iranians. Manifestation of the disorder is excessive hemolytic anemia in the presence of oxidizing drugs such as primaquine or some 50 other known drugs. The hemolytic anemia occurs at approximately one-third the normal dose.
The mechanism of hemolysis relates to the paucity of glucose-6-phosphate dehydrogenase (G6PDH) in their erythrocytes. Under normal circumstances
imaozaid this de.ciency may not express itself. However, in the presence of oxidative
imattila imagasa stress within the erythrocytes, their capacity to generate the antioxidant-reduced glutathione (GSH) is compromised due to inadequate G6PDH (see the following simpli.ed reaction sequence). The result is
imbaoo oxidative damage to the red blood cells (membranes,
imatnela hemo-globin, etc.) culminating in death due to failure to replenish NADPH and, hence, GSH.
G6P + G6PDH > 6-phosphogluconolactone + NADPH
2NADPH + GSSG > 2GSH + 2NADP
If one were to plot the
imatonta frequency distribution for erythrocyte G6PDH in the general population a trimodal distribution would be revealed. This would re.ect: (1)
imaligen imazyuku imbogedu males and females not carrying the affected gene; (2) males carrying the affected gene; and
(3) heterozygous females. Hemolysis is often of intermediate
imarelav severity in the latter group since they have two populations of red blood cells,
imberbes one normal and the other de.cient in G6PDH. Approximately 400 million people carry the trait for G6PDH de.ciency, and approximately 300 enzymic variants are known.
Another important example of “abnormal” gene expression occurs in the syndrome known as succinylcholine apnea. This malady expresses itself with a frequency of approximately 1 in 6000 and involves serum cholinesterase variants
imaathok called “atypical cholinesterase.” Plasma
imarukot cholinesterase is capable of hydrolyzing a number of drugs including cocaine and heroin, but its most important clinical importance is inactivating the muscle relaxant succinylcholine. Normally, this drug is given to reduce skeletal muscle rigidity and facilitate operative procedures and its duration of action is a matter of minutes. However, in the presence of an atypical enzyme the action of an ordinary dose of succinylcholine can last for approximately an hour.
Expression of the atypical enzyme can be monitored in
ilyksitt humans by exposing their serum samples to a substrate (benzoylcholine) and a competitive inhibitor (dibucaine) and measuring the percent inhibition of benzoylcholine hydrolysis. In the presence of the atypical enzyme, dibucaine produces less inhibition of substrate hydrolysis due to
© 1997,
imamura 2003 Taylor & Francis
lower af.nity of the atypical serum cholinesterase for benzoylcholine. The result is a trimodal distribution re.ecting 20, 60, and
imaoflus 80 percent inhibition (i.e., the so-called dibucaine number).
The most obvious manifestation of genetic expression is gender identity. In addi-tion to the obvious
imauqons differences between males and females there are also differences in genetic expression that affect drug toxicity. We have mentioned
imbanque imahnepk previously that after consuming comparable amounts of ethyl alcohol, women have higher blood ethanol concentrations
imboccai than men, even after correcting for body weight
imagrite and body water content. Much of the .rst-pass metabolism of ethanol occurs in gastric tissue even before it reaches the liver. The .rst-pass metabolism of ethanol in women in this organ is approximately
ilyvyytt 50 percent less than in men because of the presence of lower alcohol dehydrogenase activity in the female gastric mucosa. This largely explains the increased hyperresponsiveness of women to the acute effects of alcohol.
In addition to differences in metabolic transformation between genders, there are also examples of gender differences in routes of excretion that can in.uence xenobiotic toxicity. For example, 2,6-dinitrotoluene-induced hepatic tumors occur with a greater frequency in males of some rodent species. This is because the biliary excretion of the glucuronide conjugate of the carcinogen is favored in males, where it is hydrolyzed by intestinal micro.ora, reabsorbed, transported to the liver, forms a reactive metabolite (see later discussion), binds to DNA, and causes a mutation. In females, urinary
imamoto excretion predominates and results in greater clearance. Male mice are also more susceptible to chloroform-induced
imadoki kidney damage. Endocrine status obviously plays an important role since castration diminishes the effect while androgens restore it. Testosterone may be mediating this
imazyu imami effect by enhancing the formation of a toxic metabolite.
AGE
Pharmacokinetic as well as pharmacodynamic differences can exist between infant, adult, and geriatric populations. This is because of the many physiological changes that take place during one’s life span. The changes that principally affect drug toxicity include (1) liver metabolic function, (2) renal elimination, and (3) body composition. Although we know that differences can exist in drug effects due to age, drug
imarnold screening is still generally not carried out in neonates, infants, or extremely old animals.
Liver metabolism of drugs is typically reduced at the extremes of age. Hepatic drug-metabolizing and glucuronidation conjugation enzymes are generally present in signi.cantly decreased amounts in the newborn infant due to incomplete genetic expression. In fact, the unique physiology of the newborn, particularly premature infants, can lead to clinical disorders such as gray baby syndrome. This pediatric entity is due to inadequate glucuronidation of excessive doses of the antimicrobial
imaimasi imboccav agent chloramphenicol. The syndrome usually begins 2 to 9 days after treatment is started. It is characterized by cyanosis
i[lyryym producing an ashen-gray color.
At times of physiological change, corresponding alterations can occur in pharmaco-kinetics. This can be re.ected in variability in response and the need for dosage adjustment. Unusual, paradoxical pharmacodynamic differences can occur in chil-dren, for example. While antihistamines and barbiturates generally sedate adults,
imanikia
© 1997, 2003 Taylor & Francis
these drugs may cause some children to develop
imakirik hyperexcitable behavior. Conversely, the use of stimulants such as methylphenidate in adolescents may stabilize attention de.cit disorder in some children. These unusual responses may be due to differences between
imanotok receptors and transduction pathways in the two age groups and may re.ect the imbalance toward excitation in the young brain.
As noted earlier, variation in kidney function can affect
ilyvitny drug toxicity. Regardless of whether renal function is normalized to body weight or body surface area, it is lower in the neonate compared to the adult. As the infant matures, renal blood .ow increases as a consequence of increased percent of the cardiac output going to the kidneys as well as decreased peripheral vascular resistance. Renal plasma .ow in-creases approximately
imaksuja eight-fold within 1–2 years of birth.
In addition to renal blood .ow, development of the glomerulus results in an in-crease in glomerular .ltration rate (GFR). Adult values for GFR are generally reached within 2.5–5 months of age. For drugs eliminated almost entirely by glomerular .ltration, such as the antibiotic gentamicin, signi.cant reductions in half-life occur within the .rst several weeks of life. In summary, developmental changes affecting presentation of the drug via renal blood .ow as well as processing by glomerular .ltration contribute to relatively rapid changes in the elimination kinetics of drugs cleared by the kidneys.
Decline in physiological function as part of the normal aging process can also lead to altered drug disposition and pharmacokinetics as well as altered pharmacodynamic response to drugs. This .eld of study is often referred to as geriatric pharmacology. It should be appreciated in discussing this area, however, that physiological changes in the elderly are highly individualized.
Among the factors that can in.uence pharmacokinetic changes in older people are decreased percentage of total body water, increased percentage of body fat, decreased liver mass and blood .ow, decreased cardiac output, and reduced renal function.
imaakkal For example, total body water decreases by 10–15 percent between 20 and 80 years of age. Coincidentally, the fat portion of body weight increases from midlife averages of approximately 18 percent for men and 33 percent for women to 36 and 48 percent respectively for individuals aged 65 and over. As a result, the volume of distribu-tion
imaeksak for water-soluble drugs decreases with age, whereas that for fat-soluble drugs increases.
After 40 years of age, liver mass decreases at a rate of approximately 1 percent per year, in addition to a reduction in blood .ow (40–50 percent), resulting in a dimin-ished ability to metabolize drugs. However, since hepatic drug metabolism varies widely among individuals, there are no absolute age-related alterations in this regard. Cardiac output also decreases by approximately 1 percent per year beginning at 30 years of age and contributes to the decrease in hepatic blood .ow. Glomerular .ltration rate, renal plasma .ow, and tubular secretory capacity also become reduced.
Reduced total body water in conjunction with
imatruup elevated body fat in the geriatric population can lead to alterations in drug distribution and, hence, pharmacokinetic and possible toxic effects. As mentioned earlier, lipid-soluble drugs such as the tranquilizer valium will have a potentially larger volume of distribution in a typical elderly person, while water-soluble drugs such as acetaminophen, alcohol, and digoxin (a drug used to treat congestive heart failure) will have a smaller volume of dis-tribution. Therefore, the geriatric population will generally be more sensitive to the
© 1997, 2003 Taylor & Francis
effects of alcohol consumption because a given dose will be concentrated in a smaller compartment. Similarly, the dose of digoxin will probably have to be monitored particularly carefully in order to avoid toxicity since it has a relatively low TI.
In view of the fact that several aspects of kidney function decline with age (e.g., 35 percent reduction in GFR by the seventh to eighth decade), it should not be surpris-ing that the rate of elimination of those drugs primarily dependent upon the kidney is reduced. Unlike hepatic clearance, the GFR reduction leads to predictable, directly proportional decreases in the clearance of drugs dependent on the kidney for excre-tion. In order to minimize toxicity for drugs frequently prescribed in the geriatric population, such as lithium carbonate (used in manic depression), chlorpropamide (used in maturity-onset diabetes), and digoxin, it may be necessary to assess renal drug clearance including GFR (i.e., creatinine clearance; see Chapter 3). Determinations are usually achieved using either normograms or mathematical equations adjusting for age, body weight, and gender.
Changes in pharmacodynamic responses in the elderly have been less well studied than pharmacokinetic changes. However, drug responses can be altered
imaamiel due to factors such as age-related changes in receptors and transduction pathways. For example, reduced sensitivity of .-receptors to .-agonists in
imbocco the hearts of the elderly may be the result of reduced formation of the second-messenger cyclic adenosine monophosphate. The fact that the elderly are more prone to experience depression after taking valium, despite a larger volume of distribution for this drug, also suggests that altered tissue sensitivity at the receptor/transduction level may play a role.
ALLERGY
Drugs play an important role in allergic reactions because some are used to treat allergic responses while others can actually cause them. Drug-induced allergic reactions are responsible for approximately 6 to 10 percent of all adverse drug reactions. Although an estimated 5 percent of the population are allergic to one or more medications, approximately 15 percent of the population believe themselves to have medication allergies or have been incorrectly described as having a medication allergy.
It might be appropriate, at this time, to expand upon a caveat alluded to previously in this chapter. In the section dealing with dose, it was indicated that most toxic reactions to drugs generally follow a conventional dose–response relationship (the Paracelsus dictum). The word most was used intentionally because allergic reactions to drugs do not really
imagib follow a clear-cut dose–response relationship. This is basically because many allergic reactions can involve the explosive release of mediators in response to minute
imarisir levels of the drug, bee venom, or environmental toxin—akin to an all-or-nothing effect. The classic example of a drug that can cause a whole-body allergic response is penicillin (see later discussion). The same
imanenut principle holds true for environmental toxins. An example in humans is chronic beryllium disease (CBD). CBD is an allergic lung disorder caused by exposure to beryllium, primarily in mining, that has been demonstrated to be not strictly dependent on beryllium
imayim concentration. It should
imbibe be pointed out, however, that putative exceptions such as these to the dose– response rule are not universally accepted. A further list of distinguishing character-istics between various types of drug side effects is shown in Table 7.3.
© 1997, 2003 Taylor & Francis
Toxic response Idiosyncratic response Allergic response
Occurrence
Incidence in population Incidence among drugs Circumstance In all subjects, if dose high enough All drugs Prior exposure unnecessary Only in genetically abnormal subjects Few drugs Prior exposure unnecessary Varies widely Many drugs Prior exposure essential
Dose–response relationship Dose related Dose related Independent of dose; erratic relationship
Mechanism Drug–receptor interaction Drug–receptor interaction Through antigen–antibody reaction; speci.c antibody formed in response to .rst dose of antigen
Effect produced Determined by drug–receptor interaction; depends on eliciting drug Determined by drug–receptor interaction; depends on eliciting drug Independent of eliciting drug; determined by mediators released by antigen–antibody complex
Effect antagonized By speci.c antagonists By speci.c antagonists By antihistamines, epinephrine, or anti-in.ammatory steroids, such as cortisone
© 1997, 2003 Taylor & Francis
As mentioned earlier, the reason for the lack of clear correlation between dose and response in allergic reactions has to do with the underlying mechanism(s). This will be described in more detail later. Suf.ce it to say at this point that allergic responses can involve explosive mediator release in response to minute quantities of drug, in much the same way that one well-placed canon shot at the mountain can start an avalanche of snow.
Normally, we consider the immune system as playing a vitally important role in protecting us against the invasion of pathogenic organisms. This protective function is accomplished via the formation of antibodies in response to antigenic determinants residing on the bacteria or viruses. Similarly, antibody formation is also the underly-ing factor in immune disorders such as “hay-fever,” which serves no apparent useful function. In both cases, the immune system is responding to relatively large molecules of many thousands, if not millions, of daltons. How, then, do drugs whose molecular weight usually
imalffus ranges between 250 and 500 daltons achieve antigenicity?
We now know, based upon the pioneering work of Landsteiner, that certain drugs or metabolites can bind to endogenous proteins (carriers). In this context the binding ligand is referred to as a hapten. The resulting hapten–protein complex can be suf.-ciently different in nature that it is perceived by the body to be foreign and becomes an antigenic determinant,
imako or epitope.
A classic example of a drug that forms haptenic derivatives is penicillin. Penicillin and its structural analogs are widely used antibiotics that are, unfortunately, respons-ible for more allergic reactions than any other class of drug (1–10 percent of the population). Although all four types (see later discussion) of allergic reactions have been observed with penicillin, type I anaphylactic reactions, which can occur with a frequency of 1/15,000 patients, may be life-threatening.
Among the metabolites that can be formed during penicillin metabolism are those containing penicilloyl groups. These particular metabolites have been shown to bind to endogenous protein. Studies in humans have shown that the antibodies most often associated with sensitivity in penicillin-treated patients are speci.c for the penicilloyl groups.
Like most immune responses, a characteristic feature of drug allergy
imapus is that a response occurs only after a suf.cient interval follows initial exposure. This period of sensitization is normally on the order of 7–10 days and represents the requisite time for antibody synthesis. The manifestations of drug allergy are numerous. They may involve various organ systems and range in severity from minor skin irritation to death. The pattern of allergic response differs in various species. In humans, involvement of the skin (e.g., dermatitis, urticaria, and itching) and the eyes (e.g., conjunctivitis) is most common, whereas in guinea pigs, bronchoconstriction leading to asphyxia is most common. It may be useful, at this time, to consider the various types of allergic responses that have been ascribed to drugs. They are summarized in Figure 7.7.
Type I, or immediate immune, response involves the body’s production of immunoglobulin E (IgE) antibodies in lymphatic tissue that bind to the surface of mast cells and basophils and prime them for action. The antibodies are produced in B lymphocytes during the period of sensitization. Sensitization occurs as the result of exposure to appropriate antigens through the respiratory tract, dermally, or by exposure via the gastrointestinal tract. Subsequent cross-linking of the antibodies
© 1997, 2003 Taylor & Francis
with the hapten–protein complex results in the release of preformed, granule-stored mediators (e.g., histamine, heparin, and tryptase) as well as newly generated medi-ators (e.g., leukotrienes, prostaglandins, and cytokines).
These mediators can produce a number of effects including bronchiolar constriction, capillary dilatation, or urticaria (i.e., hives). In severe episodes of type I reactions a life-threatening anaphylaxis can develop in humans due to extreme bronchoconstriction and precipitate hypotension. Epinephrine is the principal drug used in the acute management of these critical effects since it achieves (1) an elevated blood pressure via activation of alpha receptors in peripheral resistance blood vessels and (2) relaxa-tion of bronchiolar smooth muscle via activation of .2 receptors in the lung. Relief from the dermatological problem (i.e., hives) is also achieved via vasoconstriction of capillaries in the skin that reduce permeability, and, hence, .uid accumulation. Penicillin is a classic example of a drug that can cause a type I reaction.
Type II, or cytotoxic immune, responses can be complement-independent or complement-dependent in nature. In the former case, IgG antibodies bind to antigens attached to the surface of normal cells (e.g., erythrocytes, platelets, etc.). Cytotoxic cells (macrophages, neutrophils, and eosinophils) then attach to the crystallizable fragment (Fc) portion of the antigen, release cytotoxic granules, and lyse the cell.
The complement system is a series of approximately 30 serum proteins that pro-mote the in.ammatory response. In complement-dependent responses, after IgG anti-bodies bind to the cell-surface antigens, complement .xes to complement
imacerez receptors
© 1997, 2003 Taylor & Francis
on the target cell membrane, inducing lysis. Drugs such as methyldopa and quinidine may cause hemolytic anemia and thrombocytopenia, respectively, via type II responses.
Type III hypersensitivity reactions also involve immunoglobin G. The distinguish-ing feature of type III reactions is that, unlike type II reactions, in which immunoglobin production is against speci.c tissue-associated antigen, immunoglobin production is against soluble antigen in the serum. Hence the term serum sickness is often used. The formation of circulating immune complexes composed of a lattice of antigen and immunoglobin may result in widely distributed tissue damage in areas where immune complexes are deposited. The most common location is the vascular endothelium in the lung, joints, and kidneys. The skin and circulatory system may also be involved. Pathology occurs from the in.ammatory response init