Report ON AN EXPERIMENTAL AND CLINICAL COMPARISON OF THE THERAPEUTIC PROPERTIES OF DIFFERENT PREPARATIONS OF 914 (NEOSALVARSAN)
EARLY in 1915 the Medical Research Council were requested by the Board of Trade to undertake the biological testing of the different preparations corresponding to 606 (salvarsan) and 914 (neosalvarsan), which were to be offered for sale and for therapeutic use in this country. When 606 was first introduced, Ehrlich drew attention to the chemically uncontrollable variations in the toxicity of the product, and the necessity for testing every batch, to ensure absence of .undue toxicity, before
... oxicity, before it-was issued for use. Every batch of salvarsan made in Germany has, accordingly, from the first been tested at the Georg-Speyer Haus, in Frankfurt, before its issue. -There is no evidence that the same practice has been adopted in the case of neosalvarsan ; the pamphlet accompanying each package of this product contains no indication of the performance of a biological test, such as is described in detail in the pamphlet accompanying each ampoule of salvarsan. In the case of neither product is there any indication of a control for therapeutic efficacy ; the control described is purely for freedom from undue toxicity, and that only in the case of the older 606. This relatively simple procedure may have been adequate in the circumstances of production in Germany, where the whole supply of both preparations has been made in one factory, working in close touch with the inventors. The Problem in England. The problem presented in this country was somewhat different. The directions for making 606 were fairly complete in the patent specifications, and the only task before the manufacturers concerned was to find those precise working details, of the somewhat complicated and uncertain process, which would give, with reasonable uniformity, a product corresponding not only in chemical composition, but in physical properties with the original salvarsan, and having no greater direct toxicity than this preparation, as I revealed by tests on animals. This was achieved by several manufacturers, and a satisfactory supply of < 606 was obtained, was supplied throughout the war, and is still available. The position was not so clear in the case of 914. The details given in the patent specification are not adequate, and the composition of the product differs widely from that suggested by the theoretical formula. A substance corresponding to this formula would 1 contain much more arsenic than the neosalvarsan to which the clinical worker was accustomed, and could not i safely be supplied for use in the same dosage. Each i manufacturer, therefore, had to supplement by his 1 own ingenuity the imperfect information available, < in the endeavour to make a product which could safely and adequately replace the familiar neosalvarc o u l d san. In these circumstances it did not seem justifiable ' to follow what was, apparently, the German practice j i with regard to neosalvarsan, and allow these substitutes to be issued without biological control. A new basis of testing had, therefore, to be formulated, founded not upon any existing German precedent, but upon what could be discovered from the literature and from actual tests of such samples of the original preparation as could be obtained. After a preliminary experimental period, which was prolonged by various conditions then existing, the basis was adopted on which products of this type have now long been and still are regularly tested. It was decided that only such batches should be allowed to be issued for therapeutic use as were tolerated by mice, on intravenous injection, in a dose of 0-3 mg.-per gramme of bodyweight. Each batch is tested on five mice in this dose, and if more than one mouse dies, or if definite bad symptoms are produced, the batch is rejected. Recent experience with an adequate series of batches of German neosalvarsan shows that the dose had been correctly chosen for the limit of tolerance for this product; nearly every batch of original neosalvarsan which has thus been tested will pass the test on the dose of 0-3 mg. per gramme of mouse, but all would fail if the dose were raised to 0-4 mg. per gramme. Two Distinct Classes of Neosalvarsan Type. A short experience with the application of this test sufficed to indicate that the preparations of the neosalvarsan type submitted for test fell into two wellmarked classes. On the one hand were those which resembled the German product more or less closely in appearance, in solubility, and in the rapidity with which their solutions underwent decomposition if left standing exposed to the air. On the other hand, there were others which had an advantage over the German product in their extremely free and rapid solubility; and which gave evidence of much greater stability in solution. Preparations of the former class had a border-line toxicity from the point of view of the test; many samples just passed it, and many others failed. A slight raising of the standard would have excluded nearly all. Those of the latter class-the freely and immediately soluble type-passed the routine test with an almost unbroken regularity ; when, experimentally, sample batches of this type were tested on higher doses, it was found that many were tolerated in a dose of 0-5 mg., some even in a dose of 0-6 mg. per gramme ; i.e., the toxicity was often only onehalf of that at which they would still have passed the official control. No question of a control of therapeutic efficacy arose at this time. It would have been impossible, in any case, to carry it out with the staff available under war conditions. From the clinical side there was no hint of dissatisfaction ; here again, in the hurry to. get men through their treatment and return them to duty, no adequate control of results was possible. The practitioner appreciated the additional convenience afforded by rapid and perfect solubility, and the freedom from constitutional reactions, even when. highly concentrated solutions were injected. The demand for the more soluble and less toxic type ofproduct grew very rapidly, until manufacturers whose endeavour had been rather to copy the German product found themselves forced to modify their process, so as to produce a more soluble type. Therewith the toxicity of their products fell likewise to the lowerlevel. When, under peace-time conditions, it became possible to follow more thoroughly the effects of treatment, it was reported from several sources to the-Salvarsan Committee (appointed meanwhile by the Medical Research Council) that the curative action of the 914 products in use, which by that time had almost entirely conformed to the more soluble and stable type, was under serious suspicion of beinginadequate. An investigation was therefore under-taken by those charged by the Medical Research Council with the duty of testing these remedies. Theywere able to use, with slight modifications, .the' method, -then recently recommended by American workers, of evaluating the therapeutic power on micf infected with a strain of trypanosomes. They found 780 DR. H. H. DALE AND OTHERS: EXPERIMENTS WITH M14. indications of a pronounced inferiority in therapeutic action of the 914 products of this country, as compared with the original German product, or with the former output of some British manufacturers. The facts were fully presented to the manufacturers concerned ; the latter, with their experience, were then able to produce preparations which, while retaining some of the advantages of rapid solubility, showed as good therapeutic qualities as the German preparation, according to the experimental indication, and still satisfied the official control as regards toxicity. These are the British products of the 914 type which are now being supplied. The Need for Official Control of Products of the 914 Type. The result of this investigation made obvious the desirability of imposing, if possible, an official control of products of the 914 type for therapeutic potency, in addition to the existing one for absence of abnormal toxicity. The existence of the latter alone, in the case of a substance of ill-defined composition, such as 914, tempted the manufacturer to secure a smooth passage for his product, by slight changes in the process, which gave it regularly a lower toxicity than that which the test allowed. It was clearly desirable to make certain that, in so doing, he was not weakening its therapeutic action. Before steps in this direction could be considered, however, it was necessary to ascertain whether the difference in efficacy of different samples in removing trypanosomes from the peripheral circulation in mice corresponded with a clearly recognisable difference in their efficacy in removing the spironemes from syphilitic lesions in the human patient. The following sections of this paper describe the details of the investigation, in which the same representative samples of the different types of product were compared both experimentally and clinically. It will be seen that, so far as the clinical observations permit conclusions as to relative therapeutic efficacy in syphilis, the results of the clinical trial are in full accord with the experimental estimates. The experimental method used for estimation of the relative therapeutic value of different preparations of 914 is a modification of that recently described by Voegtlin and Smith (1920).1 These workers determined the least dose of a preparation requisite to produce a certain effect on rats in which trypanosomiasis is developing. By counting the number of organisms pcr c.mm. in the tail blood of infected rats, Voegtlin and Smith were able to select animals suffering from the disease to the same extent, and were then able to compare the effect of different doses of the curative preparation. In our experiments the first results were obtained by the use of rats infected with Trypanosoma rhodesiense, and we observed, as did Voegtlin and Smith, the effect of a dose of 914 on the number of trypanosomes in the peripheral blood at the end of 24 hours. Our 'supply of rats ran short, however, and, as we were accustomed to use mice for the biological test for the toxicity of samples of 914, we determined to attempt to carry out the therapeutic test also on mice. About the same time, through the kindness of Prof. Mesnil of the Pasteur Institute, we obtained a strain of Tequiperdum, and we infected mice with this species. We found mice much more convenient than rats for the therapeutic test ; they were much easier to handle, and the procedure of intravenous injection into a large number of animals was much less exhausting. We made a further departure from the details 1 C. Voegtlin and H. Smith : Journ. Pharm. Exper. Ther.' 1920, xv., 175. described by Voegtlin and Smith in observing the effect of the curative dose until 72 hours had elapsed from the injection, for we found this to be a more suitable end-point than the shorter period of 24 hours. While the bulk of' our work has been carried out on mice infected with T. equiperdum, observing the curative effect for a period of three days, it should be stated that we obtained closely similar results, using rats infected with T. rhodesiense and observing the therapeutic effect only for 24 hours. It will be sufficient to outline our precedure in the case of the experiments with mice infected with T. equiperdum. When the strain of this organism was first received, it was transferred to rats and passed through several animals of this species in order that a constant virulence might be obtained. The strain has been transferred from rat to rat ever since, and when we desired to carry out a therapeutic test, mice were infected from the blood of a rat by injecting into them intraperitoneally 0-1 c.em. of an emulsion of the infected rat's blood in 1 per cent. citrate saline, such that the total number of trypanosomes received by the mouse was in. the neighbourhood of 7-8 million?. Method of Experiment. For one experiment, as a rule, 30 mice were in. fected in this way, and two days later an attempt was made to select from the 30 mice those in which the infection had progressed to a sufficiently uniform stage. This was done by pricking the tail of each mouse and drawing up a drop of blood into a hsemacytometer pipette exactly as in carrying out a redcell count. As a diluent, Toison's fluid was used, to which a little formalin had been added. When the blood had been thoroughly mixed with the stain, a drop from the pipette was transferred to the counting chamber in the usual way, and after being covered with a cover-slip was allowed to stand for five minutes in order that the organisms present might settle down on to the floor of the chamber. These were seen, under the 1/6th mm. objective, to be sharply fixed and stained, so that their enumeration was as easy as that of the leucocytes. In our experience two days after infection, out of 30 mice about 20 contained trypancsomes in the peripheral blood in numbers varying from 100,000 to 500,000 per c.mm. Voegtlin and Smith found that in rats, for the purpose of com. paring the therapeutic effect of doses of different sizes, animals whose blood contained 100,000 to 300,000 organisms per c.mm. could be regarded as being at a uniform stage of infection. A series of determinations showed us that in mice a dose producing a certain degree of curative action on animals containing 100,000 trypanosomes per c.mm. could exert the same effect on those containing as many as 500,000 per c.mm., but not on those in which trypanosomiasis had developed further, unless the dose were relatively large. Mice, in which the number of organisms lay within the limits 100,000 to 500,000 per c.mm., have, therefore, been regarded as infected to a " standard " extent, and the use of mice with slightly larger numbers of organisms has been limited to the study of the effect of doses considerably greater than what we describe as the minimal curative dose. ' The various samples of 914 preparations examined have in every case been administered intravenously, a procedure very simple to carry out in mice bv injection into one of the tail veins. The effect of the injection has been observed by microscopic examination for the presence of trypanosomes in drops of blood taken from the tail and examined fresh under 1/6th mm. objective. Such examinations of the blood were made 24, 48, and 72 hours after the injection of the drug. A certain convention was adopted in recording the results of the examination. In the fresh, living condition, trypanosomes are very easily detected, on account of the disturbance of red corpuscles due to their lashing movements. If systematic search through a drop of the fresh blood failed to reveal the presence of any trypanosomes. the result was recorded as " nil." This does not 781 DR. H. H. DALE AND OTHERS : EXPERIMENTS WITH 914. mean that we supposed that the animal had been rendered perfectly free from trypanosomes ; but, as a matter of practical observation, we found that, at the end of 72 hours, the trypanosomes were either so abundant that they were readily found in a drop of blood, or so scarce that none were found. In the former case, the infection always showed a rapid re-acceleration of its progress, and the animal died TABLE I.-Experimental Examination of Sample C of British Manufacture. (k) = :BIou<ie. (B)=Size of infection before injection, in thousands per c.mm. (C)=Dose of Cl calculated in mg. per g'. (D) =1B0. of trypauosomos present in peripheral blood after periods of hours. D=dead. __- _. -__-_ within the next two days. In the latter case, the trypanosomes were not discovered by the routine examination of one drop of blood for several days afterwards, and were usually not found for upwards of a week, the animal living for a fortnight or more longer. Although, therefore, the finding " nil " does not represent an absolute cure, it does represent a sharply differentiated and easily recognised degree of therapeutic action. It was not difficult to determine the minimum dose which would regularly produce this effect. Throughout this investigation, the dose, determined for each preparation, has been the least .dose which caused, within 72 hours, a disappearance, TABLE III.-First Comparison of British and German 9 J Products. '(1)) -" Preparatioll. (Mi.) minimal curative dose. (Ma.)= Maximum tolerated dose. (1tD)=Ratio of curative dose to tolerated dose. I complete in the above sense, of trypanosomes from the tail blood of a mouse with a standard injection. Such a dose has been termed the minimal curative dose. Description of Tables. We may first give in full detail the results of the examination by this method of two preparations, one of which is the German product, and the other a preparation of different make, purporting also to be 914, but showing a marked inferiority in therapeutic value when examined by the test. Each table is a ' , compilation of data obtained in three experiments 'on three sets of infected mice. The second column I in the tables gives the number of trypanosomes in each e.mm. of the tail blood of the mouse, which was found shortly before the injection was made. The third column expresses the dose administered in terms of mg. per gramme of body-weight; while, so far as possible, mice were chosen of 20 grammes weight, the dose was always calculated so as to be proportional to their exact weight (to the nearest gramme). The last three columns give the result of the examination of a drop of fresh blood at the stated intervals after the injection. Roughly, the result recorded as tttt means that trypanosomes were present in the tail blood in numbers exceeding 500,000 perc.mm.; TABLE II.-Examination of Sample Al, which was Current Neosalvarsan of German Manufacture.