Principles of Toxicology

Toxicology may be defined as the study of the adverse effects of chemicals on living organisms. Its historical origins may be traced to the time when our prehistoric ancestors attempted to eat a variety of substances in order to obtain adequate food. By observing which substances could satisfy hunger without producing illness or death, ancient people developed dietary habits which allowed for the survival and growth of the species. In its modern context, toxicology draws heavily on knowledge in chemical and biological fields and seeks a detailed understanding of toxic effects. Much of toxicology today involves studies of the effects of specific substances on specific biological and chemical mechanisms.

One of the fundamental concepts of toxicology is that the dose determines the toxicity. As noted by Paracelsus (1493–1541), “All substances are poisons; there is none which is not a poison. The right dose differentiates the poison from a remedy.” Thus, the answer to the question, “Is this substance toxic?” must always be, “Yes, if taken in a large enough dose.” Thus, two of the primary objectives of toxicology are to quantitate and to interpret the toxicity of substances.

I Dose–Response

Since there are both toxic and nontoxic doses for any substance, we may also inquire about the effects of intermediate doses. In fact, the intensity of biological response is proportional to the dose of the substance to which the organism is subjected. Thus, as the dose of a substance approaches the toxic level, there is no one point at which all of the organisms in the group will suddenly develop toxic symptoms. Instead, there will be a range of doses to which individuals in the test group respond in similar ways.

Once the response has been properly defined, information from dose–response experiments can be presented in several ways. A frequency–response plot (Figure 1.1) is generated by plotting the percentage of individuals with a specific defined response as a function of the dose. If a range of doses of a particular hypertensive agent is administered to a group of patients, there will be a certain number of low doses where none of the patients will yield a specific response, which in this example could be a blood pressure of 140/100. The highest of these doses without the response is the “no-observed-effect level” (NOEL) indicated in Figure 1.1. As the dose is increased, the percentage of individuals responding with the 140/100 blood pressure will increase until a dose group where the maximum number of individuals within the group responds with this blood pressure. This dose, determined statistically, is the mean dose for eliciting the defined response in the population under study. As the concentration or dose of the hypothetical hypertensive agent is further increased, the individuals previously responding with the defined blood pressure will develop yet higher blood pressures. Eventually a dose will be reached at which all the patients within the test population respond with blood pressures higher than the defined level.

The curve that is generated by these data has the form of the normal Gaussian distribution, and therefore the data are subject to the statistical laws of such distributions. In this model, the numbers of individuals on either side of the mean are equal and the area under the curve represents the total population of individuals tested. The area under the curve bounded by lines from the inflection points (indicated A and A′) include the number of individuals responding to the mean dose plus or minus 2 SD from the mean dose, or 95.5% of the population. This mean value is useful in specifying a dose range over which most individuals respond in the same way.

Frequency–response curves may be generated from any set of toxicological data where a quantifiable response is measured simply by recording the percentage of subjects that responded at each dose minus the percentage that responded at the lower dose. Generally, the frequency–response curve obtained by experiment only approaches the shape of a true Gaussian distribution. Such curves illustrate clearly, however, that there is a mean dose where the greatest percentage of individuals will respond in a specific way. There will always be individuals who require either greater or smaller doses than the mean to elicit the same response. Individuals responding to smaller doses are called hypersensitive and individuals responding to greater doses are called hyposensitive.

Dose–response data and, in particular, information concerning the acute toxicity of substances are often presented as cumulative response vs. dose. In this case, various doses of the substance are administered to groups of individuals and the percentage of individuals responding in a specific way is noted. In the case of a lethal response, the number of individuals that died is noted. In the case of a nonterminal response, such as modification of blood pressure, the number of individuals responding in each group with at least a certain specified blood pressure is noted. Prior experiments establish the broad range of doses over which the response of interest occurs. Data are plotted as a cumulative percentage of individuals responding in the desired manner vs. dose (Figure 1.1). Again, a range of doses too small to elicit a response is administered to establish the NOEL. As the dose increases, the percentage of responding individuals in each test group continues to increase until a dose is reached beyond which 100% of the individuals in the test group will respond.

The shapes of the cumulative–response curves are generally sigmoidal with a nearly linear portion at intermediate dose ranges. The mean toxic dose, or in the case of the lethal effect, the LD50, is established from such curves (Figure 1.2). The LD1 and LD99 are determined in like manner. The slopes of the linear portions of these curves for different substances need not be in the same and the relative toxicities of these substances depend on the dose. These hypothetical compounds have the same LD80 and, therefore, have the same level of toxicity at this dose. However, below this dose compound A produces greater percentages of effect than compound B and in this dose range, therefore, is more toxic than compound B. At higher doses, however, compound B produces higher percentages of toxicity and, therefore, is more toxic than compound A. Based on LD50 information alone, compound A is more toxic than compound B. In comparing the toxicity of two substances the toxic response must be clearly defined, the dose range of toxicity must be indicated, and the slopes of the dose response curves must be compared.

The LD50 is a statistically determined value and represents the best estimation of the dose required to produce death in 50% of the organisms tested. The LD50 value should consequently always be accompanied by some means of estimating the error in the value. The probability range, or p value, which is most commonly used, is generally accepted to be less than 0.05. This value indicates that the same LD50 value would be obtained in 95 out of 100 repetitions of the experiment.

Although every substance will exhibit a lethal dose–response curve, there are wide differences in the LD50’s for various substances. For example, the LD50 of caffeine is estimated to be about 200 mg/kg body weight; the LD50 of botulinum toxin, one of the most toxic substances known, is estimated to be about 100 ng/kg. On the other side of the scale, the LD50 of sodium chloride is estimated to be about 40 g/kg. As a general rule, substances with LD50’s of 1 mg or less are considered extremely toxic. Substances with LD50’s in the range of 1–50 mg/kg are in the highly toxic range. Moderate toxicity is ascribed to substances with LD50’s in the range of 50–500 mg/kg. Substances with higher LD50’s than this are generally considered to be nontoxic since relatively large amounts of material must be consumed in order to produce toxicity. For example, a substance with an LD50 of 2 g/kg requires consumption of about one cup of the material to produce toxicity in an adult human. On the other hand, extremely toxic substances with LD50’s in the 1 mg/kg range require consumption of only drops to produce toxicity in an adult human.

II Safety

Safety is defined as freedom from danger, injury, or damage. Absolute safety of a substance cannot be proven since proof of safety is based on negative evidence, that is, the lack of harm or the lack of damage caused by the substance. A large number of experiments can be run that may build confidence but still not prove the safety of a particular substance. Statistically, there is always the chance that the next experiment might show that the substance is unsafe. Our concept of safety has evolved over the years, and initially a substance was probably considered safe if it could be consumed without causing immediate death or acute injury. Our knowledge of toxic effects and our ability to test them have increased to the point where we now consider a substance relatively safe if it causes no adverse effects on specific biological systems. Today, certain substances are considered unsafe (or at least suspect) if they do nothing more than cause...