Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
PANEL DISCUSSION ON ANTlMALARIALS R. C. Elderfield, Moderator Columbia University New York, New York
166 THE BIOLOGICAL BASIS FOR ANTIMALARIAL TESTS Richard J. Porter Department of Tropical Diseases School of Public Health Ann Arbor, Michigan The present paper is an attempt to give the fundamental biological background for the chemotherapeutic papers which follow. These papers will deal with the question: "How active is a given compound as an antimalarial" Before considering this quantitative question, however, we must ask a more fundamental one, namely: "What kind of antimalarial action does the drug exhibit?" It is obvious, of course, that the type of malaria must be specified. More important than this, however, we must indicate the stage of development of a particular malaria, for, unlike the bacteria, viruses, rickettsiae and most of the protozoa, the malarial parasites have an exceedingly complex life-cycle, a cycle involving several stages of development which differ from one another both microscopically and physiologically. We shall be dealing with at least three kinds of antimalarial activity - prophylactic, suppressive and curative. Since each requires destruction of a different phase of development of the parasites, we must consider how all these phases fit together in the total life-cycle of the malarial organisms. \ \ CRYPTOZO1C SCH1ZOGONY ERYTHROCYT1C SCH1ZOGONY \ SEXUAL REPRODUCTlON Figure 1
167 Figure 1 represents the general life-cycle of members of the genus Plasmodium as exhibited by those species which have been most thoroughly studied. Infection of the vertebrate host begins with inoculation of sporozoites from the salivary glands of an infected mosquito. These sporozoites invade certain cells of the tissues of the vertebrate and initiate a process of growth and multiple division known as cryptozoic schizogony. In bird malarias the parasites of this phase, known as cryptozoites, are found in macrophages, the large cells of the skin, liver, spleen, etc. , which take in and destroy foreign particles I. In malarias of monkeys and man the cryptozoic schizogony is said to occur in the hepatic parenchyma, the liver cells properZ, although it should be pointed out that there is disagreement on this point, investigators claiming either that the evidence is not yet conclusive or that at least some of the cryptozoites are in macrophages3. In those malarias which have been best studied, there are one or two generations of cryptozoic schizogony, occupying two to ten days, depending on the species. At the end of this time the young parasites produced give rise to two new kinds of parasites. One kind, designated phanerozoites . is similar to the cryptozoites. In the avian malarias they occur in macrophages or the endothelium lining most blood vessels. In the primate malarias, those of monkeys and man, they are said to develop in the liver cells. The phanerozoic schizogony is self-perpetuating. It persists throughout the total duration of the vertebrate infection, serving as a reservoir of parasites which may give rise, at times, to the erythrocytic parasites discussed below. There is evidence that cryptozoites and phanerozoites are similar not only in appearance and cellular localization but also in the physiological attributes concerned with drug susceptibility. The two cycles together, constituting those stages in the vertebrate which occur outside of erythrocytes, are commonly called exoerythrocytic schizogony. It is proposed here that the parasites of this phase be designated by the etymologically dubious but convenient term exerythro- zoites. The first one or two generations, occurring during the incubation period before the blood infections become established, are the cryptozoites (pre-erythrocytic stages of some authors). Later generations are the phanerozoites. Simultaneously with the phanerozoites there arises also from the cryptozoites a type of parasite undergoing schizogony in the red blood cells. These are the classical parasites causing clinical malaria. They are commonly known as erythrocytic schizonts, but it is suggest- ed that the term erythrozoites would be more convenient. Their reproduction is not self limited. They may be transferred to a new host by blood inoculation and will continue to reproduce there. However, in time, acquired immunity of the host brings an end to the acute attack of malaria by suppressing them to levels at which they may be detected with difficulty or not at all. At least in the primate malarias, relapse occurs by reinitiation of the erythrocytic cycle from the persist- ing phanerozoites. In at least some avian malarias erythrozoites may give rise to phanerozoites. In the primate malarias there is indirect evidence that this does not occur. The final stage in the vertebrate host is the gametocyte, a sexual parasite arising in red blood cells from erythrozoites and perhaps also from phanerozoites. Once mature, the gameto- cytes do not develop further in the vertebrate host. Ingested by a mosquito they initiate the sexual phase of the cycle, which results eventually in the production of sporozoites infective for the vertebrate. Before describing the test procedures for the various kinds of antimalarial activity it may be useful to review the essentials of the above life cycle as seen in the groups of parasites which interest us especially. The most widely used avian malarias are Plasmodium gallinaceum of chicks, P. lophurae of ducks, chicks and turkeys and £. cathemerium of canaries. They have two generations of cryptozoites, occupying two to three days. Phanerozoites and erythrozoites appear at the end of this period, become abundant, and are then suppressed to quite low levels, where they remain for months or years. Erythrozoites may arise from the phanerozoites and, at least in P.. gallinaceum and P. cathemerium. may in turn produce phanerozoites. In P. vivax of man and P. cvnomolgi of monkeys incomplete evidence indicates that there are one or two generations of cryptozoites, lasting about nine days. Erythrozoites and phanero- zoites arise together at the end of this time. The former become abundant, causing acute malaria, and are then eliminated from the blood. The persisting phanerozoites often repopulate the blood with erythrozoites at intervals of one to many months, initiating relapses. Erythro- zoites apparently cannot produce phanerozoites.
168 In P_. falciparum of man the cryptozoites occupy about the first seven days of the infection They then give rise to erythrozoites, but apparently not to phanerozoites. When the erythrocytic schizogony is suppressed, the infection terminates. We may now consider the various kinds of antimalarial activity. Four are of major importance -prophylactic, suppressive, curative and gametocidal. A drug showing prophylactic activity prevents establishment of infection in the vertebrate host. It is clear from Figure 1 and the above discussion that prophylactic action depends on destruction of sporozoites or cryptozoites or both. Tests for such activity vary in detail but show a similar pattern . The test drug is administered to animals or human volunteers during the incubation period after inoculation of sporozoites. Determination of drug activity is based on whether infection develops in the test hosts. There arises immediately the question of interpretation of what is usually called partial prophylactic action, namely a delay in the appearance of blood infection. Clearly a drug which merely delays infection has no practical value in the control of malaria. Such delay may result from at least three mechanisms. First, the drug may persist in the host long enough after admin- istration to exhibit action against the erythrozoites. Obviously such action must be ruled out. and usually experiments can be so designed that it is not troublesome Second, the drug may merely inhibit development of the cryptozoites. Finally, the drug may destroy some but not all of the pre-erythrocytic stages. Either of these last possibilities is of interest if it expresses practical potentialities. Drugs which show complete prophylactic activity at adequate dosage are partially effective in smaller doses. It is not known which of the two mechanisms is responsible. In any event, partial activity is of interest since it suggests follow-up studies which may reveal compounds of practical value. It has been pointed out that prophylactic action may be directed against either sporozoites or cryptozoites. Either possibility would be useful, and in general experiments are not designed to distinguish between them. However, the best indications are that the known prophylactic drugs act on cryptozoites. No drug has been shown to be specifically active against the sporozoites injected by the mosquito. There are two general expressions of suppressive activity. The first is clinical cure of acute malaria without eradication of infection. The second is prevention of acute malaria without prevention of infection. Both involve inhibition or destruction of erythrozoites. This is the action of quinine, quinacrine or chloroquine when given to a patient with acute clinical malaria or when administered continuously to exposed individuals. It is clear that drugs effective only against the erythrozoites cannot cure an infection, such as P. vivax. in which phanerozoites are present. Such drugs cure P. falciparum, however, for this infection does not exhibit persisting exerythro- zoites. Tests for suppressive activity in the avian malarias are simplified in accordance with the principle that we are interested merely in the effect of drugs on the erythrozoites. Birds inoculated with parasitized blood are treated with test drugs during the first few days of infection. Comparison of their course of infection with those of controls reveal action of the drugs on the erythrozoites. In the human infections suppressive activity is studied in tests imitating the clinical treatment of acute malaria. Effective drugs ameliorate or cure the malarial attack. Curative activity is shown by complete termination of infection in the vertebrate host. In the case of P. falciparum this requires merely the destruction of all erythrozoites. In the other malarias, in which phanerozoites may initiate relapses, drugs can cure infection only if they destroy the phanerozoites or both the phanerozoites and erythrozoites. Experiments designed to detect curative drug action obviously must utilize mosquito-induced infections. Effectiveness of the test drugs may be demonstrated by various methods of determining the absence of infection in the treated hosts. However, the ultimate test for curative activity is failure of the treated infections to relapse. Finally, gametocidal action of drugs is determined by the fate of gametocytes in treated infections induced by either blood parasites or sporozoites. While elimination of gametocytes has theoretical value in malaria control, since it reduces the reservoir of infection, this action is not at present considered of much practical significance.
169 Some, at least, of the above types of drug activity are qualitatively different. For instance, there is a host of suppressive drugs, such as quinine, quinacrine and chloroquine, which have no prophylactic, curative nor gametocidal action. On the other hand, the 8-amino-quinolines, such as pentaquine, are prophylactic and curative for P. vivax yet exhibit very slight suppressive action in this infection. An extreme case is seen in the esters of para-guanyl benzoic acid, which are partially prophylactic in avian malarias yet have no detectable suppressive activity. There are indications that common physiological actions are involved in prophylactic, curative and gametocidal effects, but the number of kinds of drugs known to have such actions is still too small to warrant more than a supposition. The above discussion should have made it clear that for the present, each type of drug action must be sought by appropriate specific methods. LITERATURE CITED 1. Huff. C.G. and F. Coulston, 1944. Jour. Inf. Dis. 75, 231. 1. Shortt, H. E. etjtl, 1948-1949, Brit. Med. Jour. 1948, 1225; 1949, 1006, Trans. Roy. Soc. Trop. Med. + Hyg.- 41, (6). 3. Coulston, F. 1949, Proc. Soc. Exp. Biol. + Med. 70, 360. 1 Wiselogle, F. Y. , ed. 1946, A Survey of Antimalarial Drugs, 1941-1945. \
170
171 STRUCTURE OF 4-AMINOQUINOLINES IN RELATION TO ANTIMALARIAL ACTIVITY IN AVIAN INFECTIONS Richard J. Porter Department of Tropical Diseases School of Public Health Ann Arbor, Michigan The present paper is an effort to make some generalizations concerning the relationship of chemical structure to one kind of antimalarial activity. Before examining the data we should indicate the source and some of the limitations of the information used. All the data discussed were selected from the Survey Tables'. The specific test methods are detailed in connection with those tables. While there is a great deal of other information about the compounds considered, we shall deal only with their suppressive activity against Plasmodium gallinaceum of chicks and P. loohurae in ducklings. The quantitative measures of antimalarial activity given in Tables I - V are quinine equivalents. They are obtained by dividing the dosage of quinine base producing a given degree of suppressive action by the dosage of test drug showing a comparable degree of activity. We may well question the fundamental significance of these numbers. The fact that a drug exerts antimalarial action in small dosage does not necessarily mean that it is a superior drug. Further, it should be realized that the figure represents a summation of the various factors involved in absorption, localization, metabolic alteration, excretion and specific antiparasitic actions of the compound. However, it is certainly true that, for a given host-parasite combination, differences in quinine equivalents among related compounds have some quantitative meaning. The reproducibility of the figures has been discussed by Wiselogle '. In general, two-fold differences are considered marginal. Four-fold or greater differences are definitely significant. Table I shows the effect of various terminal groups on the side-chain in chloroquine and camoquin analogs. Certain facts are apparent from examination of the aminopropylamine series, column one. The primary amine is weak, as are the long-chain groups. The moderate-sized terminal groups are very similar in their effect on quantitative activity as long as they are simple hydrocarbons. The alcohol is definitely less favorable. The remaining columns of Table I agree well with the above generalization. The last column suggests an additional principle. While the differences are rather close to the threshholds of significance, it appears that activity falls off faster in the chains with even numbers of carbon atoms than in those with odd numbers. This is reminiscent of much biochemical information on the enzymatic manipulation of fatty acids. Table II compares those analogs of chloroquine in which the intermediate portion of the side-chain is varied, keeping the terminal diethylamine. Considering tests with Plasmodium iophurae we find that the peak of activity falls in the compounds with four- or five-carbon chains. Those with two-, three- or six-carbon chains have significantly lower activity. In the compounds with branched chains activity seems to be determined rather by the length of the chain than by the total number of carbon atoms. There is a suggestion that the position of a methyl group on a butyl chain makes some difference. Hydroxyl groups do not appear to affect activity. The significance of spatial effects is emphasized by the two cyclohexyl isomers studied. Finally, the P-phenyl butyl compound and the compounds with chains interrupted by oxygen or sulfur show definitely reduced activity.
172 TABLE I EFFECT OF TERMINAL GROUP ON SIDE CHAIN Side chain in position 4 on 7-chlor-quinoline ^"2R -NH^ /OH R -NH(CH2)3Ra^ -NH(CH2)4Ra/ -NHCHMc(CH2)3R^ NH2 0. 8 4 6 NHMe 3 NMe2 15 NHEt 20 6 NHPr 15 NHisoPr 30 NEt2 30 15 istf 20*/ NHBu 10 8 NHisoBu 8 NHEtMe 15 8 NHCMe3 30 N(CH2CH2)2CH2 30 NHcyclohexyl 10 NPr2 20 NBu2 20 NAm2 3 NHex2 0. 3 0.4 NHept2 1 N0ct2 0. 3 0. 2 N(C2H4OH)2 0. 3 a. Gallinaceum quinine equivalents b. Lophurae (duck) quinine equivalents c. Chloroquine d. Camoquin
173 TABLE II EFFECT OF SIDE CHAIN -NH-R-N Et2 IN 4 POSITION ON 7-CHLOR-QUINOLINE R Lophurae Gallinaceum C length Total C (CH2)2 3 10 2 2 (CH2)3 6 30 3 3 (CH2)4 15 15 4 4 (CH2)6 3 6 6 6 CHMe(CH2)3 15 15 4 5 CHMe(CH2)4 15 15 5 6 (CH2)3CHMe 8 8 4 5 CH2CHOH CH2 3 20 3 (CH2)2CHOH(CH2)2 15 30 5 cyclohexyl 15*, 4# 20*. 40# 4 6 CH2CH Ph(CH2)2 2 8 4 (CH2)30(CH2)2 4 4 (CH2)2S(CH2)2 3 10 * 12, 108 m. p. 151 ' 14,477 m. p. 223 , tested as diphosphate 226'. The data on Plasmodium gallmaceum in Table II are less clear, although they do not disagree with the above generalizations. Table III shows a few of the many compounds in the survey tables illustrating the effects of various ring substitutions. Those selected here are the ones available in both chloroquine and camoquin series. The general agreement in order of activity between the two series is excellent. It indicates clearly that the mechanisms of antimalarial action in the two groups are related. Further, it enhances the significance of the most outstanding discrepancy, the fact that a 7-methyl group was beneficial in the camoquin series and not in the chloroquine series. It should be noted that the last seven compounds in both series are the same. However, five of the chloroquine analogs were inactive at maximum dosage, whereas all of the camoquin analogs were active. The camoquin side-chain, given various substituents para to the hydroxyl group, has antimalarial activity in itself. It is possible that the side-chain alone is responsible for the activity of the last compounds in the series.
174 TABLE III EFFECT OF RING SUBSTITUENTS ON P. GALLINACEUM QUININE EQUIVALENTS CHLOROQUINE ANALOGS CAMOQUIN ANALOGS 7 Cl 15 7 Cl 20 7 Br 15 7 Br 15 7 I 10 7 CH3 10 5,7 Cl 4 7 1 8 3CH3 7C1 4 6 OCH3 8 6 Cl 2 7 OCH3 6 6 OCH3 2 3CH3 7 Cl 6 7 OCH3 2 6,7CH3 6 7 CH3 1 (None) 3 (None) 1 6 Cl 3 6,7 CH3 0.8 5,7 Cl 3 2 ph. 0.4 8 OCH3 0.8 8 OCH3 0.4 8 Cl 0.6 8 Cl <0. 8 2 ph. 0.4 3 CH3 5 Cl <0. 8 2 ph. 7 Cl 0. 3 2 ph 7 Cl <0. 8 3 CH3 5 Cl 0. 3 5,8 Cl <0. 8 5,8 Cl 0.2 2 ph 6 OCH3 <0. 8 2 ph 6 OCH3 0.2 Table IV illustrates a fact that is often neglected in investigations such as these. Students of chemotherapy may tend to think in terms of structural formulae rather than of chemical compounds. Thus, the 6-methyl analog of camoquin shows a quinine equivalent of 10, the 7-methyl compound, 4. It might be supposed that a compound with both would be superior. That the chemistry of the compound is important is demonstrated by the fact that the composite drug exhibits activity not significantly higher than that of the weaker mono-substituted compound. Throughout the table this fact is evident. The combinations show activity in the range of that of the weaker member of the pair.
175 TABLE IV QUININE EQUIVALENTS WITH MULTIPLE RING SUBSTITUTIONS IN CAMOQUIN ANALOGS (P. gallinaceum) 6 CH3 4 7 CH3 10 Both 6 6 CH3 4 8 CH3 0. 8 Both 0.6 6 Cl 3 7 Cl 20 Both 4 6,7 Cl 4 8 Cl 0. 6 All three <0. 3 Z phenyl 0.4 7 Cl 20 Both 0. 3 3 phenyl 0. 3 7 Cl 20 Both 0.8 6 OCH3 8 7 OCH3 6 Both 2 If, despite the warning expressed above,.one were to concentrate on structural formulae, one might construct a suggestive hypothesis. It may be that in a series such as this a substitu- tion or a hydrogen atom is favorable not if it adds to the activity of a weak parent drug but if it detracts little from the activity of a hypothetical strong compound. This would explain the fact that the combination of a relatively favorable with a relatively unfavorable substitution generally gives an unfavorable result. Table V gives additional information bearing on this point, taken from the series of chloroquine analogs. The table is to be interpreted according to the following example. The second figure in column one (0. 8) is the quinine equivalent for the 6,7-dimethyl compound. The first figure in column two (1) is the equivalent of the 3-methyl compound. The second figure in column two (0.4) gives the equivalent for the combination, namely the 3, 6, 7-trimethyl compound. Table V shows that the tentative hypothesis suggested above can be held only with qualifications. The exceptions are some of the 7-chloro and 7-bromo compounds. In several cases, and especially in the 6-methoxy, 7-chloro compound, the combination is affected more by the more favorable than by the less favorable substitution. This may be explained by the possibility that the 7-chloro substitution creates a new chemical series which must be considered separately from the parent series. The examination of composites may thus reveal pharma- cologically significant chemical differences. In any event, it is obvious that the chemistry of the compounds is fundamental. One specific fact observable in Tables III and V is noteworthy. Substitutions next to the ring nitrogen, in positions 2 or 8, appear to be uniformly unfavorable. The possibilities for general conclusions from these data are limited. Their principal general significance lies in their internal self-criticism. They show what should be obvious on a priori grounds, that the superficial comparison of structural formulae has very little possibility for fundamental conclusions. It is clear that chemical as well as spatial effects determine antimalarial action. One may wish for the chemical and biochemical information necessary to a basic interpretation of the data.
176 TABLE V QUININE EQUIVALENTS WITH MULTIPLE RING SUBSTITUTIONS IN CHLOROQUINE ANALOGS (P. gallinaceam) Substituent B Substituent A (None) 3 CH3 2 phenyl 5 Cl 6 Cl 7 Cl (None) 1 1 0. 4 0. 6 2 15 6,7 CH3 0. 8 0.4 5 Cl 0. 6 <0. 6 6 Cl 2 <0. 4 7 Cl 15 4 <0. 8 4 7 Br 15 3 7 I 10 1 8 Cl <0. 2 <0. 2 <0. 4 <0. 2 6 OCH3 2 <0. 4 <0. 2 15 7 OCH3 2 <0. 2 8 OCH3 0. 4 0.2 I. Wiselogle, F Y., ed. 1946 - A Survey of Antimalarial Drugs, 1941-1945.
177 STRUCTURAL TYPES TESTED FOR CURATIVE ACTIVITY AGAINST LOPHURAE MALARIA IN THE CHICK Alexander M. Moore Parke, Davis and Company Detroit, Michigan At the war's end, emphasis in the antimalarial program at Parke, Davis and Company shifted from the development of a good suppressive drug to the search for chemical types having curative activity against vivax malaria. The new policy was implemented by a curative test, using lophurae malaria in the chick, which was developed by Dr. Paul E. Thompson and which was set up by him on a routine basis during the summer of 1948. Usefulness of the test was indicated by positive results obtained with the 8-aminoquinol ines in conjunction with quinine, by positive results obtained with the naphthoquinone, SN 12,320*, and also by negative results obtained with a number of compounds known to be devoid of curative action against vivax in man, for example, quinine, quinacrine, chlorguanide, sulfadiazine, SN6771*. and SN 8617*. By means of the new procedure, representative members of the following suppressive series were tested for curative activity: quinine analogs, 4-aminoquinolines, acridines, a-aminocresols, chlorguanide analogs, sulfonamides, dithiocarbamates, miscellaneous dyes, and other structural types. The significance of the results will be briefly discussed in terms of the structure-activity relationships within each suppressive series. Although no new series showing curative activity has yet been discovered in this program, the wide structural variations among suppressive drugs strongly suggests that other curative types eventually will be found. * Key to Survey Numbers: SN 6771 6,6'-Diallyl-a,a'-bis(diethylamino)-4,4'-bi-2-cresol. SN 861 7 4-(6-chloro-2-methoxy-9-acridylamino)-o-diethylamino-2-cresol. SN l 2, 320 2-[3-(decahydro-2-naphthyl)propyl]-3-hydroxy-l , 4-naphthoquinone.
178 DISCUSSION DR. W.C. COOPER (National Institutes of Health) discussed trials of 4-aminoquinoline compounds in man. The group provides some of the most useful iintimalarials now available. Of the more than two hundred members of the group known to be active in the avian malarias, eleven have been tried in man, and four (chloroquine, SN 10, 751 . sontochin, and oxychloroquine) have had field trials. Chloroquine and SN 10,751 have proved so far to be the best. Chloroquine has had the widest use: it combines rapid action against erythrocytic parasites, delayed elimination from the body (permitting brief therapeutic regimens or weekly suppressive dosage), absence of serious toxicity, and no acquired resistance by parasites. SN 10, 751 (amodiaquin or camoquin) appears to have analogous properties. The 4-aminoquinolines are not effective against fixed- tissue parasites and are not causal prophylactic agents. They do not produce radical cure of relapsing vivax malaria. Their mode of action is unknown. DR. FREDERICK WISELOGLE (Squibb Institute for Medical Research, New Brunswick, New Jersey): I would like permission to tell a story. During the hectic days of the malaria program, we were in the habit of sending drugs out just as fast as we could, and expecting very prompt replies from the persons who were carrying out the tests. Because Dr. Porter was so far away, we asked him, in special cases, to reply by telegram. We were particularly interested in one 4-aminoquinoline for which we had high hopes; that is, we had hopes for a quinine equivalent of 1 5 or 30, or perhaps 50. One afternoon, just after lunch, we received a telegram that this 4-aminoquinoline had a quinine equivalent of 425. We were all set to close up the office and quit work and declare the malaria problem solved. Then we began looking at that number, and were more and more intrigued as to why Dr. Porter picked such a number, which was so extraordinarily high. We were so disturbed that we finally called him up and asked him what the quinine equivalent of this compound was. He said, "Let me look it up. " He looked it up and said, "It is 4 to 5," and that is what he had told the telegraph girlf who interpreted it very literally as 4Z5. 1 would like to ask a few questions, one of Dr. Porter, or perhaps some other person who can answer it, and that is whether anything has been done on mammalian malaria which, 1 believe, was found, right at the end of the war, in a bat, and was successfully transferred to a mouse. I would like to know whether there has been any quantitative study of those compounds found active in the avian malarias. The second question I would like to ask of Dr. Moore is purely a question of information, and that is, whether, in mixing quinine with these other substances listed in his first Table at the bottom of page 1 , it is true that the tolerated dose of qu'nine was significantly lowered by the addition of these other substances. I think that is quite interesting, because I thought you gave quinine alone, 400 mgm. per kilo, then dropped down as low as 100; or is that just that you carried it down to a lower dose? DR. A.M. MOORE: The answer to the question is a technical one. The drugs were administered in the diets, and a definite percentage was mixed with each diet. The amount which a bird ate depended on how toxic the mixture was. In the case of quinine, as I recall, there was 0. 4 per cent in the diet; that is on page 2, second slide, second line down. But in the case of pamaquine plus quinine, the dose of quinine was cut to one-half the maximum tolerated dose, which would have been 0. 2 per cent in the diet. But, when pamaquine or pentaquin was added, the mixture was very, very toxic, because of the pamaquin or pentaquin, and the birds just did not eat so much as when the diet contained quinine alone. CHAIRMAN ELDERFIELD; I wonder if Dr. Fieser would care to elaborate very briefly on the naphthoquinones.
179 I might say, in arranging this program, due to the large number - I think it was some- where around sixty-eight - of chemical classes of compounds which have shown some anti- malarial activity, in order to avoid spending a week or more in discussing the relation between structure and activity, it was necessary to trim the thing down rather drastically, and a some- what arbitrary selection of groups for discussion was made, based solely on the amount of information, as far as antimalarial activity is concerned, in a given particular group. The naphthoquinones of Dr. Fieser are, I think, one of the most interesting series of substances which have been turned up. However, it seemed to us that possibly there was not quite as much information available dealing with the naphthoquinones as there was in the other groups, and that is the reason for the somewhat arbitrary choice. I see that Dr. Fieser is here and, if he would like to elaborate for a few minutes on his naphthoquinones, I am sure everybody would be glad to hear him. DR. LOUIS F. FIESER (Harvard University, Cambridge. Massachusetts): I feel like an outsider in this group, and feel rather badly about it. Some seven years ago I became very much excited about the problem of malaria chemotherapy and, as a chemist, welcomed the opportunity for active cooperation with medical scientists in advancing the whole methodology of chemo- therapeutic research. I thought that, with cooperation of chemists and pharmacologists, the standard method of analyzing and expressing bioassay data could be improved upon. I hoped that techniques could be found for promoting better absorption of orally administered drug. I thought that development of a precision method for studying the rate of metabolic drug deactivation in the naphthoquinone series might influence work on other series of antimalarials. I thought our demonstration that in the naphthoquinone series biological potency can be correlated with chemical structure only when allowance is made for varying distribution characteristics of the compounds concerned might suggest useful developments in the study of other series of antimalarials and might have some impact on chemotherapeutic research in general. I did indeed have some very able cooperation at the time from such individuals as A. P. Richardson, W. B. Wendel, and others. However, my impression tonight is that, in respect to American antimalarial research, these hopes were largely in vain. The discussions of the evening sound exactly like those that I heard in 1944. The chemist is still dutifully synthesizing compounds and nothing more; the pharmacologist is still analysing his often very good assay data by the same sloppy method of matched doses ; and the clinician is still the supreme authority who passes judgement 6f "yes" or "no" on a compound or series without reference to points of possible importance in the organic, physical, or biochemical properties of the compounds. Our attempts to develop a satisfactory antimalarial in the naphthoquinone series, continued in the face of initial failure and against the judgement and ruling of the medical directors of the CMR, and involving such unorthodox procedures as introduction into a chemical laboratory of a Warburg apparatus and a colony of infected ducks, eventually led to apparent iuccess in the form of the drug that we have called lapinone. The curative and prophylactic effect of the naphthoquinones had been well demonstrated in avian infections and here, in lapinone, was a member of the series that, according to numerous quantitative tests in normal humans, resisted metabolic degradation and afforded high blood levels of material that retained its activity for very long periods. After a long and hard battle, we were ready for another clinical trial. But we did not belong to the club. We were not working on 8-aminoquinolines or on other compounds derived from or related to quinine, atabrine, or plasmochin. We were outsiders. We lacked the right contacts. And we had to go all the way to the American University at Beirut, Lebanon, to find the active cooperation needed. There were, and are, plenty of difficulties. In this progressive Arab state malaria cases ha/e become hard to find. When my Lebanese friends succeeded in getting a few tests that proved most encouraging, American doctors said, "But what kind of malaria do they really have in Lebanon, is it really good vivax? You should get a trial against the Chesson strain; we know all about that. " Two attempts to try the Chesson strain have been made without success. In the second attempt, Dr. Martin Young in South Carolina prepared a suspension of Chesson sporozoites in serum on a certain Thursday and shipped it to us by air. My associate met the plane Thursday night, replenished the ice, took the night train to New York
180 to deliver the thermos the next day to the steward of an Arabian-American Oil Co. plane just about to leave for Saudi-Arabia. The flight went off as planned, fresh ice was obtained at Gander and at Rome, and the material was promptly transshipped to Lebanon and injected into ten pat.ents in the mental ward. But the infection did not take. Thus, we will have to try some other desperate expedient for getting evaluation of lapinone. We can, however, report that of nine patients with primary vivax infection that were given lapinone intraveneously for four days, all nine showed prompt relief of fever and parasites, and no drug symptoms, and six were without relapse after periods of from thirteen to fifteen months after treatment. It seems to me that this is a better performance, in the direction of curative action, than that of any of the nitrogen-compounds of orthodox types that I have heard discussed by the regular members of the symposium.
181 THE RELATIONS BETWEEN CHEMICAL STRUCTURE AND TOXICITY AMONG THE 8-AM1NOOU1NOLINES L. H. SCHMIDT The Christ Hospital Institute of Medical Research Cincinnati, Ohio Some of the members of this audience may wonder why a dissertation on pharmacology, and more specifically on toxicology, should be included in tonight's panel discussion on the relations between chemical structure and biological activity. The most likely reason is that toxicological studies have played a uniquely important role in the development of curative drugs belonging to the 8 -aminoquinoline series. The clinician has leaned heavily on the results of such studies in selecting new 8-aminoquinolines for clinical trial, while the chemist has used such results as a primary guide to synthesis of new compounds. There are two explanations for this unusual importance of pharmacological studies. In the first place, until quite recently, there was at hand no experimental malaria which yielded assessments of curative activity having application to the human disease. Secondly, 8-aminoquinolines as a group exhibited a variety of striking and undesirable toxic reactions, some of which precluded administration to the human subject, and most of which could be closely related to specific chemical structures. In the time allotted to this discussion, an attempt will be made to accomplish three things: first, to describe the sequence of events which led to the unorthodox practice of using the rhesus monkey as a test object for routine pharmacological studies; secondly, to describe the different reactions evoked by various 8-aminoquinolines when administered to the rhesus monkey; and finally, and principally, to set forth what appear to be firmly supported generalizations on the relationships between the chemical structures of the 8-aminoquinolines and various aspects of their toxicity. First, let us consider the sequence of events which led to the selection of the rhesus monkey as the test object for evaluating the pharmacological properties of the 8-aminoquinolines. When interest was evoked in this group of drugs, during the latter stages of World War II, data on the effects of the compounds on larger animals were meager if non-existent. Since a base-line of information was deemed essential to rational exploration of the above chemical series, it was decided to make a careful study of the reactions of various animal species to the old German drug pamaquine (Plasmochin). Supplies of a close relative, Plasmocid, were also at hand; for this reason the projected study was expanded to include work with this latter compound. As will be apparent from the discussion which follows, the fortuitous availability of Plasmocid was a remarkably fortunate circumstance. Before proceeding to a discussion of the experimental observations, attention should be called to the chemical structures of pamaquine and Plasmocid (Chart I). These compounds have CHART I HNCH2CH2CH2N(C2H5)2 CH3O Plasmocid CH3 HNCHCH2CH2CH2N(C2H5)2 CH3O Pamaquine
182 the same quinoline nucleus, with a methoxyl group substituted at position 6 and an amino group at position 8. The side chains of the compounds have the same terminal grouping, diethyl amine. The compounds differ, however, in the alkyl groups which separate the side chain nitrogens. In the case of pamaquine, a 1-methyl-butyl group is present, while Plasmocid has an n.-propyl grouping. The studies with pamaquine and Plasmocid included work with the mouse, rat, dog, and rhesus monkey. Whereas in all of these animal species there were quantitative differences in the toxicity of the above quinolines, there were no qualitative distinctions in the toxemias which developed in the rodents and in the dog. The situation in the rhesus monkey was quite different (Chart II). CHART II TOXIC REACTIONS OF THE MONKEY (Macaca mulatta) TO PAMAQUINE AND PLASMOCID Pamaquine Plasmocid Leucopenia Neutropenia Anemia Bilirubinemia Methemoglobinemia Cyanosis Loss of appetite Loss of weight Lassitude Generalized muscular weakness Pathological changes involving bone marrow, liver, spleen, kidneys, heart and brain (nuclei of nerves III, IV and VI) Hyperesthesia Nystagmus Loss of pupillary reflexes Loss of vision Loss of equilibrium Dysmetria Dysbasia Dysergia Paralysisfof lower limbs Pathological changes in brain and cord; highly localized lesions involving-the nuclei of cranial nerves III, IV, VI and VIII and associated cell groups Monkeys, intoxicated with pamaquine, exhibited marked changes in the formed elements of peripheral blood, including leucopenia, with neutropenia approaching agranulocytosis , anemia, methemoglobinemia, cyanosis, severe abdominal cramping, anorexia, lassitude, and a general- ized muscular weakness. These more obvious symptoms were associated with a depletion of the myeloid elements of bone marrow and lesions of moderate severity in heart muscle, liver, spleen, and kidneys. Moderately severe lesions were present in the central nervous system in the supraoptic and paraventricular nuclei and minimal lesions in the nuclei of cranial nerves III, IV, and VI. Chart III contains an example of the peripheral blood changes which occur during pamaquine intoxication.
183 CHART III LEUCOPENIA AND NEUTROPENIA INDUCED BY PAMAQUINE Day of Treat- ment WBC per cmm. x 1000 Distribution of Leucocytes - Per Cent Neutrophils Lymphocytes Monocytes Eosinophils Basophils 0 18.5 54 38 4 2 2 5 2.6 9 81 4 1 5 9 1.35 3 91 5 1 0 13 1. 1 1 96 2 0 1 Monkeys intoxicated with Plasmocid exhibited an entirely different picture from that just described (Chart HI). In Plasmocid intoxication, there appeared a remarkable set of symptoms referable to injury to the central nervous system. These symptoms, which appeared within one to two days after first exposure to the drug, included hyperesthesia, muscular rigidity, nystagmus (either vertical, horizontal or mixed) loss of pupillary reflexes, and loss of equilibrium and ability to coordinate muscular movements. These symptoms were associated with severe and widespread yet comparatively localized lesions in the brain involving many of the principal nuclei in the proprioceptive, vestibulo-cerebellar, auditory, visual reflex, and extrapyramidal motor pathways, the olfactory areas and to a lesser degree the anterior horn cells. A list of the principal areas of the brain affected by Plasmocid is given in Chart IV. A typical area of injury is given in Chart V, the corresponding normal region being presented for comparison. CHART IV LOCATION OF LESIONS IN THE CENTRAL NERVOUS SYSTEM OF THE RHESUS MONKEY ASSOCIATED WITH FATAL PLASMOCID INTOXICATION Proprioceptive System Column of Clarke Large cells in the central gray Lateral reticular nucleus Lateral cuneate nucleus Mesencephalic V nucleus Centrum medianum nucleus Posterolateral ventral nucleus Visual Reflex System Abducens nucleus Trochlear nucleus Lateral oculomotor nucleus Central nucleus of Perlia Edinger-Westphal nucleus Interstitial nucleus of Cajal Magnocellular area of the lateral geniculate body
184 CHART IV (Cont. ) Vestibulo-cerebellar System Descending vestibular nucleus Medial vestibular nucleus Lateral vestibular nucleus Superior vestibular nucleus Nucleus intercalatus Nucleus prepositus Nucleus of Roller Interfascicular nucleus All cerebellar nuclei Nucleus ruber (both parts) Lateral ventral nucleus Auditory System Cochlear nuclei Superior olivary nucleus Trapezoid nucleus Nucleus of the lateral lemniscus Extrapyramidal Motor System Substantia nigra Subthalamic nucleus Some of the pontile nuclei Globus pallidus (in part) Large cells of the caudate and putamen Anterior ventral nucleus Olfactory Areas Lateral mammillary nucleus Ansapeduncular nucleus Magnocellular preoptic nucleus Lateral habenular nucleus Other Areas Some anterior horn cells Large cells of reticular formation Supraoptic and para ventricular nuclei CHART V v . ' ' *SSiS . --vS -.'3$ .rr#? WMRQK? ⢠,'. ; ».V " .- Lateral oculomotor nucleus, left, and medial longitudinal fasciculus, right, monkey #1126 control. Paraffin, Morgan's stain; X 160
185 CHART V (Cont. ) Lateral oculomotor nucleus, left, and medial longitudinal fasciculus, right, monkey #1194, after 12 days' treatment with Plasmocid. Complete absence of all neuron cell bodies, glial replacement; degeneration and glial replacement in the medial longitudinal fasciculus. Paraffin, Morgan's stain; X 160. Taken from "Neutrotoxicity of the 8-aminoquinolines. I. Lesions in the central nervous system of the rhesus monkey induced by administration of Plasmocid. " by Ida G. Schmidt and L.H Schmidt. Reprinted from Journal of Neuropathology and Experimental Neurology, VII, No. 4, October 1948. By the time these observations on pamaquine and Plasmocid had been completed, nine additional 8-aminoquinolines were made available through the work of the cooperating chemists. These compounds were administered to rats, dogs, and monkeys, the same techniques being employed as in the work with pamaquine and Plasmocid. Again there were quantitative differences in the toxicity of these compounds for the rat and dog, but no qualitative distinctions. In the monkey the situation was again different (Chart VI). Certain of the compounds produced effects on peripheral blood and bone marrow, analogous to the effects of pamaquine. Others produced symptoms of central nervous system injury, comparable to reactions to Plasmocid. Still others produced cardiac arrhythmias and symptoms indicative of a postural hypotension. CHART VI REACTIONS OF RHESUS MONKEYS TO 8-SUBSTITUTED-6-METHOXY -QUINOLINES Drug Code No. 8-Substituent Nuclear Substituent Primary Effect on 1452 HN(CH2)3NH2 Heart - circulation 3114 HN(CH2)2N(C2H5)2 CNS 3115 HN(CH2)3N(C2H5)2 '
186 CHART VI (Cont. ) Drug Code No. 8 -Substituent Nuclear Substituent Primary Effect on 11889 HN(CH2)3N(C2H5)2 (5) -OCH3 CNS 5832 HN(CH2)3NM ii 971 HNC(CH2)3N(C2H5)2 CH3 Blood - bone marrow 12322 HN(CH2)6N(CH3)2 - 11191 HN(CH2)6N(C2H5)2 - 7672 HN(CH2)3S(CH2)2N(C2H5)2 " 10309 HN(CH2)21Nj Heart - circulation 11888 tr'iT/^lj \ / \ ) ii iiN^iJri2)rti \ \/ On the basis of these observations, it was decided to make routine use of the rhesus monkey as the test animal in evaluating the toxicity of 8-aminoquinolines. This decision led to studies on some one hundred forty-six 8-aminoquinoline derivatives. These compounds have been administered to more than nine hundred rhesus monkeys, in doses ranging from one- sixteenth of the lethal to twice the lethal dose. In all cases, a subacute method of studying toxicity has been employed, divided daily doses being administered for periods up to fourteen days. The results which have emerged from these studies have demonstrated remarkable, wholly unexpected and completely unexplained relations between chemical structure and pharmacological activity. Three distinctly different reaction patterns were produced by the various quinolines: (1) depression of myeloid activity of peripheral blood and bone marrow with methemoglobinemia and anemia; (2) a complex group of neurological symptoms associated with severe and widespread lesions in the spinal cord and brain stem; and (3) disturbances in the heart and circulation, associated with highly localized lesions in the dorsal motor nucleus of the vagus. Remarkable as it may seem, every compound thus far studied has produced reactions which fall primarily into one of these patterns. There are instances in which compounds which effect peripheral blood changes also exert slight effects on the heart and circulation. Generally speaking, however, there is little overlapping in the pattern of toxic reactions evoked by a given compound. The general chemidal distribution of one hundred forty of the derivatives with respect to the 8-amino substituent is shown in Chart VII. Ten of the compounds had alkyl side chains terminat- ing in a primary amino group. Sixty-four had terminal secondary amino groups. Fifty-one had terminal tertiary amino groups, while ten had the terminal nitrogen in piperidyl linkage. Within each of these major groups there were broad variations in the alkyl group which separated the side chain nitrogens. In the case of the secondary and tertiary amino derivatives there were also broad variations in the terminal alkyl grouping on the side chain and in the nuclear substit- uent.
187 CHART VII GENERAL DISTRIBUTION OF 8-AMINOQUINOLINES EXAMINED FOR THEIR EFFECTS ON THE RHESUS MONKEY No. Compounds Studied Type of 8-amino-substituent 10 -(CH2)nNH2 n = 2 to 6 64 -(CH2)nNH Alkyl n ** 2 to i 51 -(CH2)nN(Alkyl)2 n = 2 to 1 1 10 -(CH2)n Piperidyl n = 2 to 7 (piperidyl 1,2, or 4) 5 -Miscellaneous Total = 140 The work with the ten primary amines (Chart VIII) demonstrated that, irrespective of the nuclear substituent, the type of toxic reaction was determined by the number of methylene groups which separate the side chain nitrogens. When the side chain contained 2,3,4 or 5 methylene groups in straight linkage, the compounds exerted their primary effects on the heart and circulation. However, when the side chain contained 5 methylene groups in a branched chain or 6 in a straight chain primary effects were on the formed elements of peripheral blood and bone marrow. No primary amine in the group studied evoked symptoms of central nervous system injury such as those produced by Plasmocid. CHART VIII RELATIONS BETWEEN STRUCTURE AND TOXICITY AMONG 8-AMINOQUINOLINES WITH SIDE CHAINS TERMINATING IN PRIMARY AMINO GROUPS (10 compounds) HN(CH2)nNH2 (6) -OCH3 (5) -OCH3 (4) -CH3 Irrespective of nuclear substituent, type of toxic reaction is determined by number of CH2 groups separating amino groups in the side chain. (a) When the side chain contains 2, 3,4 or 5 methylene groups in a straight chain, toxic reactions involve the heart and circulation, primarily.
188 CHART VIII (Cont. } (b) When side chain contains 5 methylene groups in a branched chain, or 6 in a straight chain, toxic reactions involve the formed elements of peripheral blood and bone marrow, primarily. The studies with derivatives having secondary or tertiary terminal amino groups (Chart IX) show that the types of reactions evoked by the compound were independent of the character of the terminal alkyl substituent. Reactions were determined, however, both by the alkyl chain which separated the side chain nitrogens and by the position of the nuclear substituent. When there was a nuclear substituent at positions 7. 6, and/or 5 the type of reaction was determined solely by the methylene linkage separating side chain nitrogens. When the linkage comprised 2 or 3 methylene groups (branched or straight) the drug produced symptoms of CNS intoxication ident- ical with those of Plasmocid. When there were 4 (branched or straight) or 5 (straight) methylene* groups present, the symptoms produced were referable to effects on the heart and circulation. When the side chain nitrogens were separated by 5 methylene groups (branched) or 6 or more branched or straight, toxic reactions involved primarily the formed elements of peripheral blood and bone marrow. CHART IX RELATIONS BETWEEN STRUCTURE AND TOXICITY AMONG 8-AMINOQUINOLINES WITH SIDE CHAINS TERMINATING IN SECONDARY OR TERTIARY AMINO GROUPS (115 compounds) HN(CH2)nNHR or -NR2 (7) -CH3 -OCH3 (6) (5) (4) (2) -CH3 -CH3 -CH3 -CH3 -OCH3 -OCH3 -OCH3 -OCH3 -OH -OPh -OPh -Cl -OEtOH -NH2 -Cl -NH2 Irrespective of the nuclear substituent at positions 7, 6 or 5 the type of toxic reaction is determined by the number of CH2 groups separating amino groups in the side chain. (a) When the side chain contains 2 or 3 (branched or straight) methylene groups, toxic reactions involve the CNS, primarily. (b) When the side chain contains 4 (branched or straight) or 5 (straight) methylene groups, toxic reactions involve the heart and circulation, primarily. (c) When the side chain contains 5 methylene groups in a branched chain or 6 or more in either branched or straight chain, toxic reactions involve the formed elements of peripheral blood and bone marrow, primarily.
189 CHART IX (Cont. ) 2. The type of toxic reaction is influenced both by the nuclear substituent at positions 4 and 2 and by the number of ' llj groups separating the amino groups of the side chain. a methoxyl group is substituted at position 4, the type of toxic ion follows the pattern set by side chain variations with (a) When react derivatives having substituents at 7 6 and 5. (b) When a methyl substituent is present at position 4, toxic reactions are independent of side chain variations and involve the formed elements of peripheral blood and bone marrow, primarily. (c) When either methyl or methoxyl substituents are present at position 2, toxic reactions are independent of side chain variations and involve the formed elements of peripheral blood and bone marrow, primarily. The situation differed considerably when the 2 or 4-position of the nucleus was substituted. In that case the reactions evoked depended both upon nuclear substituent and the alkyl group separating the nitrogens of the side chain. With a methoxyl at position 4, the reactions followed the pattern set by side chain variations among compounds with nuclear substituents at positions 7, 6 and/or 5. With a methyl group at position 4, or a methyl or methoxyl group at position 2, toxic reactions were independent of side chain variations and involved the formed elements of peripheral blood and bone marrow, primarily. The final group of compounds comprises those in which the terminal nitrogen was in piperidyl linkage (Chart X). Six unsubstituted 2-piperidyl derivatives, with from 2 to 7 methyl- ene groups in the side chain and one 4-piperidyl derivative all produced effects on the heart and circulation. One nitrogen substituted derivative and one derivative in which the side chain attachment was through the piperidyl nitrogen produced effects on the central nervous system indistinguishable from those of Plasmocid. CHART X RELATIONS BETWEEN STRUCTURE AND TOXICITY AMONG 6-METHOXY-8-AMINOQUINOLINES WITH SIDE CHAINS TERMINATING IN PIPERIDYL GROUPINGS (1 0 compounds) CH Toxic reactions involve heart and circulation HN(CH2)7- ^ CH3O Toxic reactions involve heart and circulation
190 CHART X (Cont. ) HN(CH2)2 \\/ HNCH2CHOHCH2-N \ y N â V CHoUO ^ CH /n ^^ Toxic reactions involve CNS Toxic reactions involve CNS It would be satisfying if in concluding this report information could be presented would explain the radically different types of reactions evoked by 8-aminoquinolines of such slightly differing chemical structures. Unfortunately, despite substantial efforts, no tenable explanation is at hand. As has been said often, the facts are here, underlying reasons not yet forthcoming.
191 PRIMAOUINE, SN 13,272 A NEW CURATIVE AGENT IN VIVAX MALARIA: A PRELIMINARY John H. Edgcomb^ John Arnold3/ Ernest H. Yount, Jr.J/f Alf S. Alving and Lillian Eichel- berger (from the Malaria Research Unit, Department of Medicine, University of Chicago) and Geoffrey M. Jeffery, Don Eyles and Martin D. Young (Laboratory of Tropical Diseases, Micro- biological Institute, National Institutes of Health) lntroduction This paper reports the preliminary clinical experience with primaquine, 8-(4-amino- 1 - methylbutylamino)-6-methoxyquinoline, or SN 13,272, in reducing the relapse rate of the Chesson strain of vivax malaria. This compound was first tested in man early in 1948 as part of a comprehensive study of compounds related to pamaquine, using methods previously described by Alving, et al. (1948). Primaquine differs chemically from pamaquine by having a primary amine substituted for the tertiary terminal amine on the aliphatic side chain in the 8-position of the quinoline nucleus. Pamaquine. isopentaquine and the new compound, primaquine, may be considered as a family of compounds, the members of which differ only in the characteristics of the terminal amino group. Thus: CH3 JCH2CH3 HNCHCH2CH2CH2N' \;H2CH3 (Tertiary terminal amine) Pamaquine CH3 H CHCH2CH2CH2Nf ^C CK NCH3 CH3O **v^^\rfi<^ (Secondary terminal amine) Isopentaquine Reprinted from The Journal of the National Malaria Society, Vol. 9, 285-292, 1950. Additional patients have been added to those originally reported in the first table on page 192. â ' These investigations were, in part, supported by a grant-in-aid from the United States Public Health Service and, in part, carried out under contract between the Office of the Surgeon General of the United States Army and the University of Chicago. The clinical studies were also aided by the participation of Army Medical Officers assigned to the project. The studies would not have been possible except for the valuable cooperation and help given by Warden Joseph E. Ragen of Stateville Penitentiary and other administrative officials of the State of Illinois. -^Captain. MC-AUS
192 CH, CH3O HNCHCH2CH2CH2NH2 Primaquine (Primary terminal amine) Procedures and Methods In the therapeutic trials primaquine was given in the form of the diphosphate salt (56. 9% base) and doses have been calculated in terms of base weight.47 The details of the testing program have been reported elsewhere. Briefly the reneral procedure was as follows: White, healthy inmate volunteers from the general population of the Illinois State Penitentiary (Stateville Branch) who had had no previous experience with malaria were selected. A standard method of inoculation by the bites of ten infected mosquitoes was used because this gives a consistently severe infection. Drug administration was begun early in the course of the clinical attack in order to reduce acquired immunity to a minimum. During the therapeutic trials the patients were hospitalized. Parasite counts were made daily during the immediate follow-up period and at frequent intervals for periods up to one year. No case with a follow-up of less than six months has been included. It is known that 98% of all relapses, in patients infected by this technique and treated with 8-aminoquinolines , occurs before one hundred fity days (unpublished observation). Inasmuch as the variety of toxic manifestations is similar to that of pamaquine, toxicity has been expressed in terms of pamaquine equivalents. All drugs were administered orally. Choice of clinical material. All primary cases treated were characterized by having an incubation period of less than fifteen days. A limited number of patients that had relapsed after treatment with a heterogeneous group of drugs have also been included in this report, but only subjects who relapsed within thirty days after end of therapy were chosen. Craige, et al. (1947) have shown that patients with short prepa'tent or latent periods offer a severe challenge to curative drugs. Under these experimental conditions the relapse rate after treatment of primary attacks with suppresive drugs approaches 100% (Table I). TABLE I RELAPSE RATE AFTER TREATMENT OF PRIMARY ATTACKS OF CHESSON STRAIN VIVAX MALARIA (STANDARDIZED SPOROZOITE INFECTIONS) WITH SUPPRESSIVE DRUGS Drug Relapse Rate Quinine 18/18 Quinacrine 4/4 Chlorguanide 8/8 Chloroquine 8/8 Total 38/38 Primaquine was first synthesized by Dr. Robert C. Elderfield, Department of Chemistry, Columbia University, New York City. Later supplies of primaquine have been provided by Eli Lilly and Company, Indianapolis, Indiana, and by the Abbott Laboratories, North Chicago, Illinois.
193 When primaquine was given alone in six divided doses daily, curative effect was demonstrated in doses as low as 22. 5 mgm. (base) per day (Table II). TABLE II CURATIVE EFFECT OF PRIMAQUINE WHEN ADMINISTERED ALONE DURING PRIMARY ATTACKS OF VIVAX MALARIA (CHESSON STRAIN) Cases Days From End of Rx to First Relapse Follow Up (Days) Ratio: Subjects Relapsed/ Subjects Treated Symptoms MET* HGB Mean Plasma** Concentration of Primaquine F/102 + Para DAILY DOSE OF 22. 5 mgm. *** % (gamma/liter) 1. 16 15 - None 9.6 9 2. 17 14 - Abd. + 9.4 10 3. 12 13 - 4/5 None 13.9 5 4. 16 12 - None 15. 8 4 5. - - 330 Abd. ++ 8. 0 4 DAILY DOSE OF 45 mgm. *** ,. - - 292 None 20. 0 43 2. 18 16 - Abd. ++ 22.0 75 3. - - 333 1/5 None 18. 1 45 4. - - 333 Abd. + 16.9 33 5. - - 284 Abd. + 23. 8 30 * Expressed as % of total hemoglobin (average for last five days of treatment) ** Determined by method of Brodie, Udenfriend and Taggart (1947) *** Drug administered in six divided doses daily for two weeks Abd. + = mild, transient abdominal cramps Abd. ++ = moderate, repeated abdominal cramps Subsequent experience with the action of primaquine against trophozoite -induced infections suggests that many, if not all "relapses" reported in Table II were really recrudescences, that is, were due to incomplete eradication of trophozoites because the parasitemia recurred very early. When primaquine was given in conjunction with 1. 64 Cms. quinine (base) daily, consider- ably greater curative effect resulted (Tables III, IV and V).
194 TABLE III CURATIVE EFFECT OF PRIMAOUINE WHEN ADMINISTERED TOGETHER WITH 1.64 GMS. QUININE DAILY IN PRIMARY ATTACKS OF VIVAX MALARIA (CHESSON STRAIN) Cases Days From End of Rx to Follow Up (Days) Ratio: Subjects Relapsed/ Symptoms MET* HGB Mean Plasma*' Concentration of First Relapse Subjects Treated % Primaquine (gamma /liter) F/102 + Para DAILY DOSE OF 15 mgm. *** 1. 19 16 - None 4.8 13 2. 66 65 - Anorexia 15. 1 32 3. 58 57 - 4/5 None 2.2 14 4. 50 47 - None 4.9 12 5. - - 451 None 3. 3 12 DAILY DOSE OF 30 mgm. *** 1, - - 370 Abd. + 8. 7 12 2. - - 369 None 12. 1 15 3. - - 365 0/5 None 9.8 11 4. - - 365 None 11.0 14 5. - - 356 None 7. 7 13 DAILY DOSE OF 60 mgm. *** 1 . - - 403 Abd. +++ 8.2 27 2. . 405 0/4 Abd. + 6.8 58 3. - - 377 Abd. < i 9. 1 42 4. - - 367 Abd. ++ 9.2 21 * Expressed as % of total hemoglobin (average for last five days of treatment) ** Determined by method of Brodie, Udenfriend and Taggart (1947) *** Drug administered in six divided doses daily for two weeks Abd. + - mild, transient abdominal cramps Abd. ++ = moderate, repeated abdominal cramps Abd. +++ = severe, persistent abdominal cramps
195 TABLE IV CURATIVE EFFECT OF PRIMAOUINE WHEN ADMINISTERED IN DOSAGES OF 22. 5 MGM. TOGETHER WITH 1. 64 GMS. QUININE DAILY* IN PRIMARY ATTACKS OF VIVAX MALARIA (CHESSON STRAIN) Cases Days From End of Rx to First Relapse Follow Up (Days) Ratio: Subjects Relapsed/ Subjects Treated Symptoms MET** HGB % Mean Plasma*** Concentration of Primaquine (gamma/liter) F/102 + Para 1. - - 370 None 5. 3 8 2. - - 520 None 7.9 - 3. - - 365 Diarrhea 3.4 10 4. - - 415 None 6.3 8 5. . . 364 None 5.8 5 0/10 6. - - 226 None 4. 1 1 7. - - 341 None 8. 7 4 8. - - 373 None 12.2 9 9. - - 373 None 8.6 6 10. - - 389 None 10.3 4 * Both drugs administered in six divided doses daily for two weeks ** Expressed as % of total hemoglobin (average for last five days of treatment) *** Determined by method of Brodie, Udenfriend and Taggart (1947) TABLE V CURATIVE EFFECT OF PRIMAOUINE WHEN ADMINISTERED IN DOSAGES OF 22. 5 MGM. WITH 1. 64 GMS. QUININE DAILY* - IN CASES REPRESENTING THE FIRST OR SECOND RELAPSE AFTER OTHER THERAPY Cases Days From End of Rx to First Relapse Follow Up (Days) Ratio: Subjects Relapsed/ Subjects Treated Symptoms MET** HGB % Mean Plasma**' Concentration of Primaquine (gamma/liter) F/102+ Para FIRST RELAPSES 1. - - 175 None 4.6 3 2. - - 277 None 8. 3 9 3. - - 312 None 5. 7 3 4. - - 330 None 6.0 6 5. - - 337 Abd. ++ 6. 5 2
196 TABLE V (Cont. ) Cases Days From End of Rx to Follow Up (Days) Ratio: Subjects Relapsed/ Symptoms MET** HGB Mean Plasma*** Concentration of F/102 + First Relapse Para Subjects Treated % Primaquine (gamma/1 ite r) FIRST RELAPSES 6. - - 307 None 6.7 2 7. - - 364 0/13 None 3.4 4 8. - - 316 None 5. 1 3 9. - - 346 None 8.4 6 10. - - 343 None 8. 1 2 11. - - 245 None 4. 5 - 12. - - 161 Abd. ++ - - 13. - - 285 None 8. 3 4 SECOND RELAPSES 1. - - 362 None 4. 7 17 2. - - 369 Abd. t 5. 3 13 3. 76 74 - Abd. + 5.6 25 4. - - 349 1/8 None 1.9 5 5. - - 365 None 9. 3 9 6. - - 365 Anorexia 7.7 - 7. - - 349 None 9.4 5 8. - - 349 Abd. + 7.5 6 * Both drugs were administered in six divided doses daily for two weeks ** Expressed as % of total hemoglobin (average for last five days of treatment) *** Determined by method of Brodie, Udenfriend and Taggart (1947) Abd. t = mild, transient abdominal cramps Abd..f+ - moderate, repeated abdominal cramps A daily dose of 22. 5 mgm. of primaquine given concurrently with 1. 64 Gms. of quinine (base)5/prevenU-d relapse in practically 100% of cases. (Tables IV and V). Increasing the dose â» Subsequent studies have shown that 1. 64 Gms. of the base is in excess of the amount of quinine needed. A dose of 0. 82 Gms. of base (1. 0 Gms. quinine sulfate) is certainly sufficient and possibly even as little as 0. 547 Gms. of base may suffice. The smallest effective dose of quinine has yet to be determined.
197 of primaquine to 60 mgm. increased the toxicity without concurrent therapeutic advantage (Table III). Toxicity studies were carried out on volunteers unsuited for therapeutic trials. .£/ The drug was given during attacks induced by intravenous malaria. The same drug dosage regimen was followed. The toxic manifestations observed during administration of primaquine at 120 mgm. daily (alone, and in conjunction with other drugs) is shown in Table VI. The toxicity of primaquine tends to be cumulative; in some instances symptoms began late in the course of drug administra- tion and continued for several days after its discontinuance. TABLE VI TOXICITY STUDIES OF PRIMAQUINE (SN 13,272) Days Rx Case Age Weight Symptoms Laboratory Findings MET* HGB Mean Plasma** Concentration of DAILY DOSE OF 120 mgm. *** % Primaquine (gamma/lite r) 14 1. 21 145 Abd. ++ Normal 20. 5 201 14 2. 28 145 Abd. + and Nausea Normal 21.7 228 14 3. 39 168 Abd. + WC 4700**** 18. 3 308 14 4. 24 150 Abd. ++ Normal 19.6 144 14 5. 43 152 Abd. + Normal 20. 7 171 DAILY DOSE OF 120 mgm. GIVEN CONCURRENTLY WITH 1.64 gms. QUININE*** 14 I. 36 140 Abd. +++ Normal 9.0 77 14 2. 24 172 Abd. ++ + Normal 9.8 85 14 3. 28 189 Abd. ++ and Nausea Normal 8. 8 132 14 4. 39 137 Abd. + and Anorexia Normal 15. 5 126 14 5. 30 139 Abd. + and Anorexia Normal 11.1 74 14 6. 22 180 Abd. +++ and Vomiting Normal 5. -5 - * Expressed as % of total hemoglobin (average for last five days of treatment) ** Determined by method of Brodie, Udenfriend and Taggart (1947) *** All drugs given in six divided doses daily **** 5% immature granulocytes, 18% mature granulocytes, 76% lymphocytes (returned to normal five days after last dose of drug) Abd. + - mild, transient abdominal cramps Abd. ++ - moderate, repeated abdominal cramps Abd.+++ = severe, persistent abdominal cramps -> Extensive studies of toxic and therapeutic effect of primaquine in mammals and primates have been done by Dr. L. H. Schmidt (The Christ Hospital Institute for Medical Research, Cincinnati, Ohio).
198 Two hundred and forty mgm. (base) probably represents the maximum dose that can be administered with safety for periods longer than a week even under close observation in hospital (Table VII). Although toxic manifestations were severe, no irreversible damage was noted. In contrast, the maximum tolerated dose of pamaquine is probably 90 mgm. (base) per day; and, for pentaquine (SN 13,276) is 120 mgm. , but severe damage to the nervous system may result from its administration at that dose. Of the curative antimalarial drugs extensively studied, only isopentaquine (SN 13,274) can be given safely at a dose of 240 mgm. (base) per day for an extended period. It is of interest to note the effect of quinine on the production of methemoglobin. At high doses of primaquine with quinine the methemoglobin is roughly 50% as great as that formed by the same dose of primaquine given alone (Table VI). TABLE VII TOXICITY STUDIES OF PRIMAQUINE (SN 13.272) Days Rx Case Age Weight Symptoms Laboratory Findings MET* Mean Plasma** HGB Concentration of % Primaquine (gamma/liter) DAILY DOSE OF 240 mgm. GIVEN CONCURRENTLY WITH 0. 199 gms. METHYLENE BLUE*** 9 I. 24 134 Abd. ++ WC 2000* 7.5 395 DAILY DOSE OF 240 mgm. GIVEN CONCURRENTLY WITH 1.64 gms. QUININE*** 11 1. 39 160 Abd. ++ + + WC 3200## 9.6 213 14 2. 36 172 Abd. + + + WC 11,900 10.0 131 * Expressed as % of total hemoglobin (average for last five days of treatment) ** Determined by method of Brodie, Udenfriend and Taggart (1947) *** All drugs given in six divided doses daily # 13% immature granulocytes, 3% mature granulocytes, 79% lymphocytes (returned to normal seven days after last dose of drug) #<f 9% immature granulocytes, 30% mature granulocytes, 56% lymphocytes (returned to normal fourteen days after last dose of drug) Abd. ++ = moderate, repeated abdominal cramps Abd. + + + = severe, persistent abdominal cramps Abd. ++++ = intolerably severe abdominal cramps Discussion The therapeutic significance of the change in character of the terminal amino groups on the side chain of pamaquine-like compounds can be seen by the following comparison. (For greater homogeneity of data only primary cases with a standardized infection are summarized):
199 Drug* Daily Dose** (base weight) Relapse Ratio Estimated Dose for 100% Cure Pamaquine 60 mgm. "* 10/20 90-120 mgm. Isopentaquine 60 mgm. 6/20 90 mgm. Primaquine 22. 5 mgm. 0/20 22. 5 mgm. * administered concurrently with quinine ** given in six divided doses for two wecKs + this is equal to 133 mgm. of the salt, pamaquine naphthoate It is apparent that on an equal weight basis, primaquine is about four times as active as the best of the other'members of the family. Comparison of the subjective toxicity in the pamaquine family in terms of estimated pamaquine equivalents is as follows: Drug Toxicity C hem iithc rape utic Index+ In Terms of In Terms of Symptomatology at 60 mgm. /day Maximum Tolerated Dose Pamaquine 1.00 1. 00 1 Isopentaquine 0. 75 0. 33 2 1/2 Primaquine 1.00 0.33 10 + chemotherapeutic index is the ratio of largest tolerated dose divided by the smallest dose capable of preventing nearly all relapses Although not the subject of this paper, which stresses comparative curative effect and toxicity of the three drugs studied under standard conditions, it should be mentioned further that primaquine can establish a high prophylactic and curative ratio when administered in thera- peutically safe single daily doses. This is not possible with either pamaquine, pentaquine or isopentaquine. These latter three drugs have been shown in field studies to be active in doses one-half to one-third as great as those necessary to produce equivalent results against our standard test strain of vivax malaria (Most, et al. , 1946), (Alving, 1948), (Coggeshall and Rice, 1949). Observation of a limited number of patients suggests that naturally acquired infections likewise can be cured with much smaller doses of primaquine than reported here. Primaquine is superior to both pamaquine and Isopentaquine because it will cure severe infections of vivax malaria in dosages that are relatively non-toxic in white subjects and because it has a wide range between the clinically effective dose and the maximum tolerated dose. ACKNOWLEDGEMENTS We wish to acknowledge the invaluable nursing and technical assistance rendered in this study by Miss Shirley Mock, Mrs. Katherine Chellew and Mrs. Lorraine Gruben, and to the many hard-working and conscientious inmate nurses and technicians assigned to the malaria project.
200 REFERENCES Alving, A. S. , Pentaquine (SN 13,276) and Isopentaquine (SN 13,274), therapeutic agents effective in reducing relapse rate in vivax malaria, Proc. 4th Int. Cong, on Trop. Med. and Mai. , 1948, 1:734-741. Alving, A. S. , Craig, B. , Jr. , Pullman, T. N. , Whorton, C. M. , Jones, R. , Jr. , and Eichelberger, L. , Procedures used at Stateville Penitentiary for the testing of potential anti- malarial agents, J. Clin. Inv. , 1948, 27, Supp. 2^:2-5. Brodie, B. B. , Udenfriend, S. , and Taggart, J. V. , The estimation of basic organic compounds in biological material. IV. Estimation by coupling with diazonium salts, J. Biol. Chem. , 1947, 168:327. Coggeshall, L. T. , and Rice, F. A. , Cure of chronic vivax malaria with pentaquine, J. Am. Med. Assn. , 1949, 139:437-438. Craige, B. , Jr., Alving, A. S. , Jones, R. , Jr., Whorton, C. M. , Pullman, T. N. , and Eichelberger, L. , The Chesson strain of plasmodium vivax malaria. II. Relationship between prepatent period, latent period and relapse rate. J. Infect. Dis. , 1947, 80:228. Most, H. , Kane, C. A. , Levietes, P. H. , London, I. M. , Schroeder, E. F. , and Hayman, J. M. , Am. J. M. Sc. , 1946, 212:550. ADDENDUM The toxicity of primaquine in darkly pigmented races and in children is unknown. Until these problems have been adequately investigated therapeutic use of the drug should be limited to adult white subjects.
201 5-ARYLOXY-2.4-DIAMINOPYRIMIDINES AS ANTIMALARIALS George H. Hitchings, Elvira A. Falco and Peter B. Russell The Wellcome Research Laboratories Tuckahoe, New York A major project in our laboratories for a number of years has been a study of the role of pyrimidine derivatives and related substances in the biosynthesis of nucleic acids . A screening test was established using Lactobacillus casei as the test object, in a manner which allows one to test for antifolic acid, anti-thymine and antipurine effects as well as for stimulatory activity. By means of reversal experiments the nature of the blocks which the inhibitors produce can be examined in more detail. For example (Figure I) 2, 4 -diamino-5-p-chlorophenoxypyrimidine had been found in the screening test to have a strong antifolic acid activity. The figure shows a more detailed study of this effect. Over a considerable range of concentration the effect of the inhibitor can be overcome completely by the addition of more folic acid. Moreover, the level of growth obtained depends rather closely on the ratio of the pyrimidine to folic acid and is nearly independ- ent of the absolute concentration. This constitutes a rather good example of competitive inhibition and probably indicates that the two substances are competing for some surface in or on the microorganism. 00224 0112 0560 J 80 14.0 PGA- my PER ml. Figure I The inhibitory effect shown by this aryloxydiaminopyrimidine is a general property of 2,4-diaminopyrimidines and condensed systems containing the diaminopyrimidine structure. 2 These all inhibit the growth of L^. casei in the folic acid system. When reversal experiments are
202 carried out one finds that these substances form a whole spectrum of inhibitors. Some of them, like the aryloxydiaminopyrimidines, are reversed only by folic acid. Others, like diaminopurine, are reversed very little by folic acid but readily by a purine, and all gradations between these two extremes can be found. In the course of our work a number of commercially available substances had been tested, and among these, both paludrine and quinine have effects in the Lactobacillus casei test which resemble those of the diaminopyrimidines. Paludrine (chlorguanide) lies quite far toward the antifolic end of the spectrum, quinine somewhat nearer the center, that is, the inhibition by quinine can be relieved to a considerable extent by the addition of purines. This was one reason for testing the compounds for antimalarial activity. A second reason was that there is a certain structural resemblance between chlorguanide and the aryloxypyrimidine derivatives. This is shown on the slide (Figure II). If chlorguanide is written in a cyclic form, you will see that the resemblance between it (IV) and the p-chloro- phenoxy derivative (V) is close, and perhaps closer to that of the £-chlorophenoxy-6-methyl- pyrimidine (VI). This type of structural resemblance - as between a cyclic and an acyclic compound - probably is not very good chemistry, but sometimes results in leads to the solution of practical problems. Figure II At any rate, these pyrimidines were submitted for trial as antimalarials and both found to be active, V being about as active as quinine and VI about four times as active. So we were introduced to antimalarial research by the back door, as it were. L.Ooking through the literature one finds that a considerable amount of work on pyrimidine derivatives as antimalarials was carried out during the wartime search for such substances, but all these pyrimidines contained basic side chains attached to the pyrimidine nucleus. With the wisdom of hindsight, it is now possible to see that these basic side chains are not only not necessary, but definitely reduce the activity. This may be illustrated by the substances shown
203 on the next slide (Figure III). VII is one of the substances studied by the I. C. 1. group3. On their scale the 2-j>.chloroanilino-4-diethylaminoethylamino-5-phenoxypyrimidine shows t activity at 160 mg./kg. and was negative at 80 mg. per kg. At about the same time Todd and co-workers* reported on compound VIII which had an amino group in place of the anilino group in the 2-position. The activity of this was rated a + at 120 mg. per kg. and a t at 60 mg. per kg. In our own work, the 2,4-diamino-5-phenoxypyrimidine (the corresponding substance with removal of the basic side chain) would rate at least +++ on the I. C. I. scale and is thus more active than either of the more complex substances, although it is not now considered to be a very active substance. Figure III For the benefit of the chemists the next two slides show general methods for the prepara- tion of the aryloxypyrimidines. Figure IV shows the preparation of the 6-unsubstituted variety by formylation of aryloxyacetic esters,, condensation with guanidine, followed by chlorination and amination. The 6-alkyl derivatives are prepared from acyl acetic esters (illustrated on Figure V by acetoacetic ester) by way of the o-aryloxyacylacetic ester and then by the above route to the diaminopy rimidine.
204 "" - Figure IV Figure V
.205 The elaboration of this lead toward new antimalarials is being pursued as a cooperative project between the Wellcome Research Laboratories, Tuckahoe, and the Wellcome Laboratories for Tropical Medicine in London under Brigadier J. S. K. Boyd. At just about the time our work began, Plasmodium berghei infection in the mouse became available through the courtesy of Professor Shortt of the London School of Tropical Medicine and Goodwin5 studied the action of several antimalarial drugs in parallel against P_. gallinaceum in chicks and P. berghei in mice and found some interesting differences (Figure VI). The dose of quinine required to suppress the parasites is much the same on the two organisms, and chloroquine and our 2, 4-diamino-5-£- chlorophenoxy-6-methylpyrimidine have about the same quinine coefficients against both experi- mental malarials. On the other hand, mepacrine is more active against P-. berghei than against P_. gallinaceum while chlorguanide and especially pamaquin are less active against P^ berghei Our compounds all have been tested against both types of experimental malarias, and we hope that eventually data bearing on the suitability of P_. berghei as a screening test will be available. It may be that it will be found to be the mammalian malaria, suitable for large scale testing and investigation, which has been so much desired. FIGURE VI Drug P. gallinaceum P.. berghei 48-210 4. 3 4. 7 Mepacrine 2. 0 7. 8 Pamaquine 18. 0 <2 Chloroquine 13. 13 Chlorguanide = Paludrine 12. 4.6 Goodwin, L G. , Nature, 164, 1133 (1949) The elaboration of .the lead provided by the antimalarial activity of 2, 4-diamino-5-p- -hlorophenoxypy rimidine and its 6-methyl derivative has followed more or less straightforward .ines. The biological work is still rather fragmentary and it would be premature to attempt to correlate biological activity and chemical activity on a comprehensive scale. However, a few points can be illustrated. In the first place, not all of the substances of the diaminopyrimidine spectrum of inhibitors possess antimalarial activity, but many of the more potent compounds have some degree of antimalarial activity. This is true of diaminopteridines^ which lie at the antifolic end of the spectrum and of diaminopurine which lies at the opposite end. These have in common the diaminopyrimidine structure and some sort of substituent at the 5-position of the pyrimidine ring. A number of subseries are under investigation, and some of these have shown compounds of very bigh activities. The 5-aryloxy-2, 4-diaminopyrimidines? are used for purposes of illustration. (Figure VII). In the first place, the unsubstituted phenoxy derivative is not very active, and has activities of the same order of magnitude on the two test organisms. Substitution of the aromatic nucleus by halogen in the para position increases the activity somewhat but the activity against P. berghei is enhanced more than that against £. gallinaceum. The meta chloro derivative is sven more active against P. gallinaceum but considerably less active against P. berghei.
206 FIGURE VII 5 -PHENOXY -2, 4-DIAMINOPYRIMIDINES Activity and Substitution in the Aromatic Nucleus Substituent of Phenyl 6 Substituent Quinine Equivalent P. jgallinaceum P. berghei none H 0. 5 0. 3 E-ci H 0.8 1.0 m-ci H 1. 1 0. 5 none CH3 0. 5 0. 1 £-Cl CH3 4. 5 4. 7 m-Cl CH3 6.8 0. 5 £-OCH3 H >0. 5 0.0 The addition of a methyl group in the 6-position of the pyrimidine diminishes the activity against Plasmodium berghei when the benzene ring is unsubstituted but increases the activity of the p-chlorophenoxy compound by a factor of about 5. The corresponding meta derivative is even more active against P_. gallinaceum but of relatively low potency against FV berghei. One might guess from this that the electron distribution in the aromatic nucleus influences the activity markedly, and that Pâ. berghei is more sensitive to these effects than P__. gallinaceum. This supposition is borne out by all the results through several series of compounds and in general the activity against .P_. berghei is enhanced by electron attractive groupings in the para position. One further illustration of this may be found on the final line of this slide. The p_-methoxy derivative, containing an electron donor group, retains activity against P_. gallinaceum but is essentially inactive against P_. berghei. The effect of the alkyl group in the 6-position is somewhat unpredict- able. Moreover, the basicity of the whole molecule, as measured by the pKa of the monohydro- chloride, seems to be of secondary importance. Undoubtedly, steric factors are involved which may be unravelled by further work. Finally, I should like to return to the question, which was implied in the beginning, of the mechanism of action of these substances. Is their antimalarial action an expression of anti- folic acid activity? First, some indirect evidence may be derived by comparing antifolic activity in the Lactobacillus casei system with antimalarial activity (Figure VIII). The next slide shows the antifolic acid activities of some of the diaminopyrimidines and their antimalarial activities, with paludrine as a reference compound. The activity in the L. casei system is given as the amount causing 5U per cent inhibition in the usual folic acid growth medium and the figures given, therefore, are inversely proportional to the activity. It is seen that all these pyrimidine derivatives are much more active than paludrine against L.. casei but less active than paludrine against malaria. Furthermore, some related pyrimidines, kindly furnished by Professor Todd, show no activity in the L. casei system but have appreciable activities as antimalarials. Thus it would appear that if the antimalarial activity is related to the antifolic activity the nature of this relationship is somewhat obscure. Nevertheless the antimalarial activity of the pyrimidines is at least partially reversed by folic acid in vivo and potentiated by sulfonamides, just as is that of chlorguanide. (We are indebted to Dr. Joseph Greenberg for this information). Furthermore, Dr. Greenberg and his colleagues, Miss Taylor and Dr. Josephson, have found the pyrimidines to be relatively inactive in vitro. In this respect also they resemble chlorguanide and differ from other antimalarials. These facts would seem to suggest the same mode of action for the
207 pyrimidines as for chlorguanide. A direct test of this possibility on a chlorguanide-res istent strain of Plasmodium eallinaceum (again by Dr. Greenberg and associates) revealed that no cross resistance developed. Thus their action appears to be similar, but not identical to that of chlorguanide. Figure VIII Before concluding, I should point out that no work with our pyrimidines in the human malarias has been reported yet, and very little work with sporozoite transmitted infections has been carried out. Some compounds with extremely high potencies, by present-day standards, have been found, but high potency in a substance is no guarantee that it will have activity on the exoerythrocytic forms of the parasites, and no real advance would have been made in the absence of this type of activity. The outlines of a pattern relating to biological activity and chemical structure in the pyrimidine series are beginning to appear but considerably more work will be required before the whole picture can be filled in.
208 REFERENCES 1. Hitchings, G. H. , Elion, G. B. . Falco. E. A. , Russell. P. B. . Sherwood. M. and VanderWerff, H. . J. Biol. Chem. 183. 1 (19S0). 2. Hitchings. G. H. . Elion. G. B. . VanderWerff, H. . and Falco. £ A J. Biol. Chem. 174. 765 (1948). 3. Curd. Richardson and Rose. J. Chem. Soc. 308 (1946). 4. Hull. Lovell. Openshaw and Todd. J. Chem. Soc. 41 (1947). 5. Goodwin. L. G. , Nature 164. 1133(1949). 6. Greenberg. J. . J. Pharm. Exp- Therap. 9J. 484 (1949). 7. Falco. E. A. . Hitchings. G. H. , Russell. P. B. and VanderWerff. H. . Nature 164. 1O7 (1949).
