Oct., 19511 BAIN 573 The Evaluation of Drugs in Man, with Special Reference to Antihistaminics BY W. A. BAIN (Presented at the meeting of the Biological Methods Groufi on Monday, March 12th, 1951) The advantages of graded or quantitative responses over “all or none” or quantal responses in determining dose - response relationships in human subjects are discussed. There are special difficulties and limitations in the evaluation of drugs in man. As an example of the application of a graded response method, the evaluation in man of various antihistaminic drugs, by measurement of the reduction of the response to intracutaneous histamine, is described. In earlier published work, which is briefly reviewed, comparisons of relative potencies, durations of action and therapeutic efficacies were made in separate sets of experiments.More recently, estimates of relative potencies and durations of action have been derived simultaneously from the same subjects; some comparisons made in this way are presented, including hitherto un- published data for chlorcyclizine, chlorprophenpyridamine and compound Antihistaminics will only reduce, and not abolish, the effects of the 3-pg dose of intracutaneous histamine used in these experiments. They may, however, prevent or abolish the symptoms of hay fever and chronic urticaria in doses that produce only a small percentage reduction in the effect of the 3-pg test dose of histamine. From a consideration of the dose - response curve to intracu taneous histaminics, an explanation for this seeming discrepancy between pharmacological and the therapeutic results is suggested, and a clearer picture of how the drugs produce their therapeutic effects is obtained..B.W. 405-C-49. THE evaluation of drugs in man involves essentially the application to man of those principles of quantitative pharmacological investigation that have hitherto been applied chiefly to biological assay and st andardisation procedures on animals and isolated animal tissues. I t was, indeed, largely for the solution of the problems of biological standardisation that these methods were first developed. In general they are methods that permit, among other things, the establishment of dosage - response relationships. The pharmacologist has two different types of data that can be used for the determination of such relationships.The first is got by measuring the degree of effect produced by a drug (graded or quantitative response), the second by noting the presence or absence of a drug effect (“all or none” or quantal response). The quantitative method is capable of yielding much more information than the other and is preferable when it can be applied. It has the further advantage that, if the same subjects are used for all the comparisons, only small numbers need be used. The quantal method, besides yielding less information, involves larger numbers of subjects. In man, however, because of subjective or other effects that can be arbitrarily graded, responses that would be quantal in animals can often be made to yield data that are essentially quantitative and can be dealt with as such.Further, there is usually more than one effect or manifestation of a particular drug that can be recognised and may be of importance in man: consequently, even when the different effects are strictly “all or none,” there is more information to be extracted from the quantal type of response in man than from the corresponding type in animals. I t seems that the future advancement both of human pharmacology itself and of rational pharmacotherapeutics must depend largely on how far such methods can be developed and applied to the investigation of drug actions in man. The extent to which they can be developed will itself depend largely on the success achieved in recognising or establishing measurable criteria of the effects of the drug under test.Within a class of drugs the short-term evaluation in man should involve not only the determination of relative weight for weight potencies when the drugs are administered by their usual routes, but also the determination of relative rates of onset and duration of action and the relative immediate toxicities. This pharmacological evaluation can sometimes be Both are important.574 BAIN: THE EVALUATION OF DRUGS Il\i MAN, WITH [Vol. 76 carried out on normal people, but at other times only on patients suffering from some particular condition. In either event, since the object of obtaining the pharmacological information is to make possible its rational application in pharrnacotherapeutics, a therapeutic evaluation of the most promising drugs of the class should !succeed the pharmacological one.Often, however, the pharmacological evaluation on patients will constitute, by its nature, the short-term therapeutic evaluation as well. Quantitative comparisons of drugs in man are subject to the same variables as are comparisons in animals and animal tissues, but additional difficulties and limiting factors immediately present themselves. Thus, in obtaining dosage - response data in animal experiments, an objective record of the whole effect of a given dose can often be obtained, leaving no doubt about the maximum effect. Even where no objective record is possible, the number of observations yielding the required data for a given dose can be made sufficiently great to give a smooth curve of response. Such curves are clearly necessary for dosage - response data in man and to determine difference:; in duration of action of different drugs, but continuous and objective records of the drug effect can seldom be obtained, and the number of observations that can be made to determine the rise and fall of the effect is limited.Further, the dose-range used in animals can be as great as desired, but is necessarily limited in man. The element of personal danger is never entirely absent in human experiments, and particular caution has to be exercised in investigating new or hitherto untried compounds. Finally, since only small numbers of subjects are usually available for such tests, all com- parisons have to be made on the same people, who, if the experiments are numerous and of long duration, begin, sooner or later, to find them exceedingly tedious.The illustrations in this paper are simple examples of the application of a graded response method to the comparison of antihistaminics in man. The antihistaminics are drugs of varied structure, whose most important action is to antagonise, presumably by receptor competition, certain of the actions of histamine. Most antihistaminks have other actions, such as the hyoscine-like effects (both central and peripheral) seen on systemic administration and the anaesthetic effects produced by local application to suitable sites and lesions. The hyoscine-like effects, while sometimes useful therapeutically, more often constitute the most undesirable side-action of the drugs. Since histamine itself, injected into the skin, gives characteristic effects-the wheal and the flare-that can be measured, the quantitative modification of either or both of these effects by antihistaminics can be used as the basis for an evaluation and comparison of different antihistaminic drugs in man.The techniques used in our original experiments have been published1*2 and need not be detailed here. Three drugs, promethazine, mepyramine and antazoline, were compared for relative potencies and durations of action, and the first two for their therapeutic efficacies. Comparisons of potency were made by determining, for each drug, under standard conditions, the maximum reduction of the intradermal histarnine response in each of a group of six subjects to three different oral doses of the drug.The average maximum response to each dose of each drug was expressed as the average percentage reduction of wheal area. As the dose-response curves for the three drugs were parallel, a simple ratio expressed the potency differences. Promethazine was about seven times more potent, weight for weight, than mepyramine and about fifteen times more potent than antazoline. Relative durations of action were determined by following the reduction of the effect of an intradermal dose of histamine, usually 3 pg, to and beyond its maximum after adminis- tering approximately equi-effective doses of the different antihistamine drugs to another group of subjects. Relative durations of action were expressed as the average times taken for the full effect of the drug to be reduced to half.For promethazine, mepyramine and antazoline these times were 194, 5Q and 34 hours respectively. A therapeutic comparison of promethazine and mepyramine was then carried out on 20 patients with chronic urticaria. All these had first been treated with mepyramine and the daily dose to control the urticaria was determined. They were then taken off all therapy and, when the reactions to intradermal histamine had returned to normal, treated with promethazine. This drug, in view of its long duration of action, was given in a single dose at night, the dose being adjusted to give as nearly as possible the same effect as the mepyramine in the earlier period of treatment. The results showed that promethazine was, on the average, about fourteen times more potent therapeutically than mepyramine.In the pharmacological tests it was about seven times more potent. It is clearly to the prolonged duration of action of promethazine that thisOct., 19511 SPECIAL REFERENCE TO ANTIHISTAMINICS 575 difference between the pharmacological and therapeutic potencies was due. A further point brought out by this comparison was that the effective dose of promethazine produced fewer side reactions than the corresponding dose of mepyramine, so that 14 of the 20 patients preferred the promethazine treatment. These first comparisons of potency and duration of action were made in separate experi- ments on different groups of people. The necessity for reducing the total number of observations to a minimum has led us t o modify this proced~re,~ and now all the necessary I I 1 1 I I t .1 L I I I I I I 1 1 I I I I I I I I 100 3 0 0 1000 3000 I0000 10 30 TIME, MINUTES Fig. 1. Graphical representation of time - response data for compound “405” and chlor- Left-hand family of curves, effects of 12.5. 25 and 50 mg of “405”; right-hand Points of maximum action and of half-action prophenpyridamine. curves, 18.5, 25 and 50 mg of chlorprophenpyridamine. are indicated on the graphs information is obtained simultaneously from the same subjects. Each subject receives duplicate or triplicate intracutaneous injections of 3-0 pg of histamine, estimated as base, in 0.05m1, and the resulting wheal areas are recorded 5 to 7 minutes later. The anti- histaminic is then taken on an empty stomach. The test dose of histamine is injected half- hourly until the maximum effect has been reached, and then hourly or at longer intervals, until the effect has obviously passed its half maximum. By plotting the percentage reduction of wheal area against time for each subject, an approximate curve of action for the dose of the drug in each subject is determined.From each such curve is derived an estimate of the maximum effect, the time for the establishment of this and the time for it to be reduced to half. The individual results in the group are averaged. Three different doses of each drug are usually investigated. Table I gives the main data abstracted from nine experiments of this kind on three different drugs. A graphic representation of the results with two of these, chlorprophen- pyridamine, and a compound prepared by Dr.Adamson of the Wellcome Laboratories and known as “405,” is shown in Fig. 1. The comparison of these drugs is of particular interest, as they both have the same potency but differ in the rate at which the maximum effect is reached and, more important, in the rate at which it falls to half. Compound “405” thus seems an exception to the general rule that the more potent an antihistaminic the longer its duration of action. The averages of the individual maximum percentage reductions of wheal area for each dose of an antihistaminic (see Table I) constitute the data from which dose -response relationships are obtained. Fig. 2 shows such relationships for some drugs, derived in this576 BAIN: THE EVALUATION OF DRUGS I N MAN, WITH TABLE I SUMMARY OF DATA ON RELATIVE POTENCIES AND DURATIONS OF ACTION FOR THREE ANTIHISTAMINICS [Vol. 76 Drug Dose, mg Chlorcyclizine 25.0 100.0 250.0 Chlorprophenpyri- 12.5 damine 25-0 50.0 “405” 12.5 25.0 50.0 Average of individual maximum percentage reductions & S.E.23.2 -k 2-4 44.3 5 2-7 57.9 * 2.3 36.0 & 4.5 48.3 i 2.1 56.7 & 1.7 39.1 2 1-75 46.6 & 2.66 55.8 & 2.36 Times to maximum and to half-maximum action 7 A 7 Average of individual half-action times, Average time i.e., time to half- for all doses Average half- action less time to maximum action time to maximum action, action for all doses 932 - 348 = 584 294 minutes 1024 minutes minutes 1082 - 278 = 804 (5 hours) (17 hours) 1940 - 255 = 1685 1600 - 190 = 1410 187 minutes 2050 minutes 2640 - 188 = 2452 (3 hours) (34 hours) 2470 - 183 = 2287 760 - 120 = 640 122 minutes 679 minutes 863 - 128 = 735 (2 hours) ( l l g hours) 779 - 117 = 662 way from a group of six people. The slopes of the lines are less than in our original experi- ments on other subjects and are such that if the dose to produce a 30 per cent.reduction in the response is taken as unity, then for a 65 per cent. reduction it is ten, and, assuming the lines to maintain their courses indefinitely, the dose to abolish the effect of 3 ,ug of histamine is approximately 100, i.e., about 1-5 g of promethazine or 10.0 g of mepyramine! (The doses of the different drugs required to produce a 30 per cent. response are indicated below the abscissae in Fig. 2 and give a measure of the relative potencies.) U W 4 J U W a $ 5 0 a a 2 t- U ’> w UI 0 4 t- z W U w a a 70 60 5 0 40 30 2 0 10 0 00 ORAL DOSE, m9 Fig.2. Dose - response relationships for different antihistaminics. The dose of each pro- ducing-a 30 per cent. ieduction in whLal areais shown below the abscissa; these indicate relitive potencies. Thus, taking promethazine as 1, the corresponding doses required to produce the same maximum effect are : chlorprophenpyridamine (and “405”) +, chlorcyclizine 24, mepyramine 6*. antazoline 144, thonzylamine 20Oct., 19511 SPECIAL REFERENCE TO ANTIHISTAMINICS 577 I t is clearly impossible to abolish the effects of a 3-pg intracutaneous test dose of histamine with any of the antihistaminics at present available. Equally clearly, it is possible to prevent or abolish many of the effects of histamine liberated in the body by doses of these drugs that produce only a moderate reduction in the response to the 3-pg intracutaneous test dose.Clinical experience, indeed, especially with the longer acting drugs, indicates that the symptoms of some allergic conditions, e.g., hay fever, can be controlled by a level of anti- histaminic effect represented by no more than a 10 or 15 per cent. reduction of the 3-pg test wheal. How, then, can these percentage reductions of wheal area be interpreted so as to give a picture of how the antihistaminics produce their therapeutic effects? 0.001 : 0.602 0 :012 0.1 : 0.2 5 1.0 i 1.5 10.0 100.0 INTRADERMAL HISTAMINE, I*g Fig. 3. Shift of dose - response curve for intracutaneous histamine by an antihistaminic.Experiment on five subjects. Assuming the regressions are parallel, the shift shown, representing a 25 per cent. reduction of the 3-pg test wheal, involves a six-fold increase in the dose required to produce any given effect. For further explanation see text Upper line before and lower line after dosage of antihistaminic. As the percentage reduction of wheal area that we measure is an expression of the shift of the dose - response curve to intracutaneous histamine, the following considerations provide a possible answer. The upper graph in Fig. 3 shows the log-dose - response regression for intracutaneous histamine in five of the subjects used in the foregoing experiments. The lower graph shows the shift of this 3 hours after ingestion of an antihistaminic.If it is assumed that the two regressions are parallel, the shift is such that a six-fold increase in dose is now required to produce the same effect as originally. Now the percentage reduction of wheal area to the 3-pg test dose in this experiment is about 25. Thus a 25 per cent. reduction of wheal area involves a six-fold increase in the dose of intracutaneous histamine required to produce any given original effect within the dose range represented on the graphs. By drawing parallel dose - response regressions through points representing different percentage reductions of the effect at the 3-pg dose level, the relationship between the percentage reduction of wheal area to this test dose, and the extent to which any dose of intracutaneous histamine must be raised to produce its original effect, is immediately established.This relationship, calculated from Fig. 3, is shown by the lower line in Fig. 4. If the slope of the dose - response curve to intracutaneous histamine is shifted slightly, so as to meet the abscissa at 040.2 instead of 0*0015pg, the upper line of Fig. 3 results. Thus, though the exact relationship will depend on the slope of the dose -response regression, it is obvious that a relatively small percentage reduction in the intracutaneous wheal response represents a considerable578 BAIN: THE EVALUATION OF DRUGS I N MAN, WITH [Vol. 76 increase in the dose of histamine required to produce any given response, including a threshold response. Thus it is not difficult to see how the effects of extrinsic histamine liberated in the body may be prevented or abolished by doses of antihistaminics that produce only a moderate reduction in the effect of an intracutaneous test dose, for this relatively small reduction may involve increasing many-fold the threshold for the production of extrinsic histaminic effects.Fig. 4. Relationship between percentage reduction of wheal area from a 3-pg test dose of histamine by an orally administered antihistaminic, and the extent t o which the intracutaneous dose of histamine must then be increased to produce the same effect as originally. Abscissa: reduction of wheal area from 3-pg intracutaneous test dose. Lower graph from data based on normal (upper) dose - response line in Fig. 3. Upper graph shows effect of modifying slope of normal dose -response line so that it cuts abscissa in Fig. 3 a t 0.0015 instead of 0.002.Note that 10 per cent. reduction of wheal area more than doubles dose required to produce a given effect. Similarly, a 30 per cent. reduction involves a 9 to 10-fold increase, a 60 per cent. reduction an 80 to 100-fold increase and so on. Ordinate : multiple of dose of intracutaneous histamine (log scale). For further explanation see text This affords one way of resolving what many clinicians regard as an impossible discrepancy between pharmacological experiments and clinical experience. In conclusion it should perhaps be pointed out that the methods very briefly outlined here, while probably capable of development to the same degree of precision as other quantita- tive biological methods, have been used simply as a better means than any other available of obtaining comparative information about antihistaminics in man.But no undue stress should be placed on the numerical results. Nor must it ever be forgotten how wide is the individual variation in response to these drugs4 and how necessary it is, in their clinical use, to take account of this. REFEREKCES 1. 2. 3. 4. SCHOOL OF MEDICINE Bain, W. A., Broadbent, J. L., and Warin, R. P., Lancet, 1949, 3, 47. Bain, W. A., PYOC. Roy. SOC. Med., 1949, 42, 615. Bain, W. A., Batty, J. E., and Broadbent, J. L., 1951, Communication to British Pharmacological Bain, W. A., Hellier, F. F., and Warin, R. P., Luncet, 1948, ii, 964. Society, Jan., 1951. DEPARTMENT OF PHARMACOLOGY UNIVERSITY OF LEEDS DISCUSSION MR.N. T. GRIDGEMAN asked if it would not be useful to take, as a measure, response (reduction of PROFESSOR BAIN replied that he thought i t better to deal with these separately. wheal area) multiplied by time from maximum to half-maximum effects.Oct., 19511 SPECIAL REFERENCE TO ANTIHISTAMINICS 579 DR. H. 0. J. COLLIER commented on the difficulties encountered in the experimental development of new paralysants because their relative potencies in man did not agree with their relative potencies in experimental animals. Did the relative potencies of antihistaminics assessed by Professor Bain’s method agree with those obtained by any method using animals ? He thought it was clear, however, that drugs that were very weak antihistaminics by animal tests proved weak in man, but drugs that were powerful in man often did not seem to be so in animal tests.But i t was obviously the results on man that were the more important. DR. L. J. HARRIS enquired whether tests had been made in which two drugs were administered simultaneously-for example, one producing a rapid, short-lived amelioration together with another producing a more prolonged action. Might the resultant time curve be expected to be a summation of the two individual curves? PROFESSOR BAIN said there were differences of opinion on this point. PROFESSOR BAIN replied that this would be expected, but had not been tried. DR. J. V. SMART asked if the magnitude of the initial response to intradermal histamine had any PROFESSOR BAIN replied that there was no apparent correlation, i.e., individual variation in the response DR. J. RAVENTOS wondered what was the explanation of the increased effect of histamine immediately PROFESSOR BAIN said he had originally attributed this to an inhibition of diamine oxidase, subsequently DR. F. J. DYER asked whether there was any difference in response of different age groups. PROFESSOR BAIN said that no clear differences in responses of different age groups had been detected. effect on the amount of an antihistaminic necessary to produce a given response. to an antihistaminic did not seem to depend on the magnitude of the intradermal histamine reaction. after the administration of promethazine that was apparent in one of Professor Bain’s graphs. masked by the antihistaminic action of the drug, but this was by no means certain. The difference between individuals was much greater.