Pharmaceuticals for Developing Countries: Conference Proceedings (1979)

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.

THE SCIENTIFIC BASE: OPPORTUNITIES FOR RESEARCH

INTRODUCTION Joshua Lederberg In his introduction, JOSHUA LEDERBERG indicated that the panel presentations are intended to elaborate rational scientific and tech- nological approaches to the health problems of developing countries. He noted that never before in history have the opportunities been so great, yet so poorly exploited for human benefit. Despite the very substantial breakthroughs in basic biological knowledge, which can be labeled as the new molecular biology, relatively little has been ex- ploited to further the understanding of host-parasite interactions. Biological specificity is the basic approach to therapy and prophylaxis. Scientists are challenged to find the means to destroy the parasite without destroying or harming the host. This concern can be traced to Paul Ehrlich and his search for "magic bullets," using the general principle of specific action as exemplified by the staining pro- perties of micro-organisms. Today, there is still much to be learned of the biochemical bases of specificity, even for the arsenicals and antimonials used extensively in the past. Currently, investigators have focused their concerns on the metabolic pathways of parasites and hosts. They have mounted a search for wedges that can be driven between these pathways toward development of effective therapies. This concern is the subject of several papers that follow. Another source of specificity, which illuminates our understanding of host-parasite interactions, are the immunoglobulins: to exploit the host capacity to protect itself against parasitic invasion. Dr. David will report about his investigations in this field. Specificity can also be examined with the tools of contemporary biology to exploit advances in genetics and molecular biology. DR. LEDERBERG indicated that his presentation would elaborate further on this approach. The range of topics for discussion could easily have encompassed other exciting fields, £.j?., studies on the morphology and life cycle of the parasite as sites for other wedges to be driven between parasite 282

and host. Thus, Sidney Brenner has pointed out how the water excretion mechanism of trematodes, embodied in a single cell, constitutes a highly specialized physiologic target for specific pharmaceutical development. 283

COMPARATIVE BIOCHEMISTRY AND THE DESIGN OF CHEMOTHERAPEUTIC AGENTS FOR INFECTIOUS DISEASE Seymour S. Cohen Introduction In an earlier paper in this Conference, Janssen described three strategies for control of infectious disease.^/ In this paper we shall discuss only one of these, that of chemotherapy for the treatment of disease, and shall focus on the problem of the design of new drugs that are disease-specific. It is assumed that: (1) new and better drugs are necessary, as stated recently in a discussion of the chemotherapy of trypanosomiasis;^/ (2) research to discover new pharmaceuticals can be carried out; and (3) new promising drugs can be tested in humans and can be delivered to needy patients. All of these assumptions have been challenged in this Conference, but it seems more useful to follow the lead of the sponsors of our meeting at this time. As the title of this paper suggests, we shall consider a general strategy in approaching chemotherapy of infectious disease. This strategy is considered applicable to all infectious disease,^/ and not only for parasitic diseases of major importance in the developing world. The strategy stems in significant measure from our perception of the history of all chemotherapy in the last 70 years and our understanding of the present state of biomedical science and technology. Briefly, the strategy rests on the following facts: 1. All infectious agents differ in major biological and chemical properties from their mammalian hosts. 2. Major chemical differences stem from differences between genetic nucleic acids of parasite and mammal. The definition of these 284

differences constitutes one advanced aspect of the study of comparative biochemistry. 3. These differences in nucleic acid composition and sequence between organisms compel differences in structures of essentially all proteins and enzymes of the parasites and their hosts. 4. The existence of specific differences between these proteins of two organisms establishes the possibility of the synthesis of agents capable of inhibiting or inactivating proteins essential to multiplica- tion of the parasite. 5. During the last decade improvements in technology of protein isolation and characterization have made definition of specific areas in a specific crucial protein of a parasite a feasible undertaking. 6. Improvements in design and synthesis of inhibitors permit exploitation of the definable differences between proteins. It is suggested then that specific chemotherapeutic agents can be designed to inactivate proteins essential to multiplication or survival of an infectious agent, whether it be bacterium, virus, protozoan, fungus, or worm. This paper will now elaborate aspects of the history of the development of these facts and indicate some particular applica- tions of the strategy to some of the major parasitic diseases in devel- oping countries. By 1907, Paul Ehrlich had observed cure of trypanosomiasis in mice by trypan red and had formulated a goal of a specific chemotherapy, _i.e_. , the selective elimination of a parasite from a diseased animal without major damage to the host._4/ His subsequent efforts towards such a goal led to development of Salvarsan for syphilis, but as we now know, the arsenical is toxic in important respects. Thus, Ehrlich's goal was not realized before 1936 with the discovery of the chemotherapeutic efficacy of Prontosil and of its essential component, the relavitely nontoxic sulfanilamide. This finding and the demonstration that sulfanilamide competes with p-aminobenzoic acid and inhibits synthesis of essential folic acid in bacteria opened the era of a potentially rational chemotherapy. At most, the compounds that were made then were designed to emulate substrates and cofactors of essential reactions at the active sites of enzymes. Many of these substances, though inhibitory, are not sufficiently selective between active sites of enzymes of the parasite and those of the host. Nor have inhibitors been tailored generally to complement essential regions other than active catalytic or regulatory sites in the crucial protein. Nevertheless, in attempts to inhibit some enzymes, of which dihydrofolate reductase is an outstanding example, this approach has 285

led to development of discriminatory chemotherapeutic agents. Agents that successfully inhibit dihydrofolate reductase include pyrimethamine, trimethoprim, and methotrexate, some of whose properties are presented in Table 1. As shown in this table, selectivity of those compounds active on dihydrofolate reductase demonstrated that enzymes of similar (analogous) function from many cells may be distinguished by their widely different reactions to various inhibitors. Thus, this table demonstrates that structures of these proteins differ sufficiently to permit development of inhibitors selective for analogous bacterial, protozoan, and mammalian enzymes. Shortly after discovery of sulfanilamide, and with the advent of World War II, work on penicillin was resumed. The development and effi- cacy of this antibiotic led to further investigations to discover other therapeutic agents produced by organisms. An antibiotic was detected initially by its antibacterial action, and the selectivity of its toxi- city was explored relatively late in the sequence of studies seeking therapeutic efficacy. This search has also produced major new substances, and creative synthetic modifications of antibiotics have been introduced, e^£., peni- cillin-1ike derivatives to improve stability, distribution, and anti- bacterial spectrum. Nevertheless, new antibiotics have been sought empirically and have generally been found to take advantage of unique structural and metabolic properties of bacteria, as compared to the mammalian hosts. In these respects, discovery of antibiotics has not only been of enormous import in the progress of therapy and in preven- tive medicine but also has demonstrated that clear structural and meta- bolic differences exist between antibiotic-sensitive organisms and their hosts. Selective Chemotherapy and Comparative Biochemistry By the 1950s, these differences were accepted as important elements of evidence of the major division between procaryotic and eucaryotic cells. Thus, penicillin had led to detection of the charac- teristic peptidoglycan in many bacterial cell walls, and streptomycin, chloramphenicol, erythromycin, and others led to demonstration of impor- tant differences between procaryotic and eucaryotic ribosomes. In the following decade, observations indicating that few serological cross reactions were detectable between the proteins of procaryotic and eucaryotic organisms began to be understood in terms of transcription and translation of specific nucleotide sequences into unique proteins. These structural specificities exist despite func- tional similarities among the proteins. Thus, an E. coli enzyme, such as the phosphofructokinase active in bacterial glycolysis, is very different from the human phosphofructokinase. Indeed, Mansour and Bueding had demonstrated twenty-five years ago that a schistosomal 286

z m _, o • tH • • M iH o 0 O H U O o H u m 03 CSl EX! z »-H CM H • 1 B O Z D. -H 3 U >> 3 T3 o 0 0 u --. M O- l-i CM 0 &S m H CU CU i- H w O. O, CO U 0 U m os rH ^ •H a! o o 0 o O CO PL, 60 %4 • • • CJL* 0 P^ o X 41 Q M .2 H CO a w s W O <U s~\ z o3 4-1 V Ct, ^ o o d CO Oi U o 0 O U Q CO O 0 •o H U 3 l-i <: H O X o 4-1 o O O J ^ i-H d fi-i O M x^ M CO CU H M Z -<J « CM H • — CO 1.4 o 0 O O < 4j 0j 0 H cd ^ o /-N U dc iH A S -""^ i- H vO 00 Q <M X ^ H d U o 0 <r> co <; m <u OO i- H 0 Z -J i > o 0 0 3 -H M Cn DC ^H o H O CO < as OS Q [-H ^- i ~~^ Z XI 03 1 ^*^ ^*^ u U M •^1 u] i- H -H U Q CU i- H CO 4-1 Z d a CU CO -H U3 O fe CU •H •H 4-1 iH M CO u o O B ^4 cd 1-1 CU i- H c 0. ||Q CO 4J CX CO n o Hl H O O * 4J si cd CO CU fH CO CU 4_1 4J B & d Xl B CU O •H -H -H u 9 •H S £ 4J 4J 4-1 CO iH 4J d d d p3 ^ ^4 <u ^J ^J ^J ^ BL, H *^^. -.*^. -*^. H cd | -Q j U 287

phosphofructokinase differed from the mammalian enzyme, and was selectively inhibitable by an appropriate antimony derivative.^,_7/ Thus, it was already known two decades ago that bacteria and other parasites contained unique essential proteins theoretically sensitive to specific inhibitors. At that time, however, it was not known how to design an inhibitor capable of reacting specifically with particular proteins. Seeking a Penicillin for Virus Infection Although the synthetic and antibiotic-screening programs of pharmaceutical laboratories generally do not deliberately take advan- tage of these opportunities to utilize specific protein inhibitors, many effective antibacterial agents have been obtained by a variety of other efforts encompassing brute screening or enlightened empiricism. However, in applying these methods to virus infections, these extra- polations have generally failed, despite improved screening procedures adapted to viruses or virus-infected cells. After 30 years of such efforts and the expenditure of large sums, the problems of selective chemotherapy presented by virus infections have for the most part not been amenable to screening procedures. For example, the importance of nucleic acid synthesis in virus multiplication led to testing many ana- logues of bases and nucleosides. With few exceptions, inhibitory com- pounds so developed proved similarly inimical to the multiplication of viruses and to growth of normal cells. Nevertheless, it is relevant to examine the sparse harvest of the few antiviral agents that do possess some selectivity. Figure 1 shows some agents active in inhibiting infections with herpes simplex virus. Compounds in Group A are activated to the nucleoside monophosphate by the unique deoxypyrimidine kinase determined by virus DNA. These acti- vated nucleotide analogues have been the true inhibitors of virus multi- plication. Compounds in Group B affect the unique virus-induced DNA polymerase. Thus, the few substances found to have promising selectivi- ty in inhibiting virus-infected cells specifically support the generali- zation described for antibacterial agents. The virus-induced thymidine kinases and DNA polymerases differ from the host enzymes, and these differences can be exploited. In fact, with the exception of viroids, all viruses code for specific enzymes and structural proteins essential to virus multipli- cation, stability, and survival. A large body of information on multi- plication of various viruses suggests identity of specific proteins induced by the viral etiological agents of herpes, influenza, foot and mouth disease, and some tumors. Many of these are essential to virus multiplication, and the essentiality of these proteins should be noted. Thus, the deoxypyrimidine kinase induced by herpes virus is not believed to be an essential enzyme, and selection of kinaseless mutants 288

in Infected cells by analogue-substrates of the kinase may lead to resistance to the drugs of Group A in Figure 1. FIGURE 1. RELATIVELY SPECIFIC INHIBITORS OF HERPES VIRUS MULTIPLICATION AlUdR HOCH2 HOCK OH OH w araT acyclo-Guo BCdR B. araA H,0,P-CHoCOOH PAA A» Compounds whose antiviral activity requires phosphorylation by the herpes-induced deoxypyrimidine kinase: 5'-amino-5-iodode-oxyuridine (AlUdR) , ara T, 9-(2-hydroxyethoxymethyl)guanine (acycloGuo) and 5-bromo- deoxycytidine (BCdr); B. Compounds that do not utilize the herpes-induced deoxypyrimi- dine kinase: ara A and phosphonoacetic acid (PAA).8/ 289

On the Technology of Protein and Enzyme Characterization Until recently, knowledge that proteins induced by etiologic agents differed from those of the host was mainly of academic interest. Protein purification and characterization were slow and difficult, requiring relatively large amounts of material. Advances in the past decade have made it possible to use relatively small amounts of organ- isms for rapid purification and characterization of proteins and enzymes. For example, as a result of such advances: 1. One gram wet weight of trypanosomes has permitted isolation and purification to homogeneity of the protozoan glutamate dehydro- genase, which was then shown to differ from the mammalian enzyme.^/ 2. Picomole quantities of myoglobin have been sequenced.10/ 3. Proteins of more than l,000 aminoacids, such as J3-galacto- sidase, have been completely sequenced.ll/ 4. Immunochemistry, crystallography, and organic chemistry have helped define sensitive regions of proteins, as have combinations of genetic and complementation techniques when applied to subunit assembly. 5. These techniques have been used to direct synthesis of poly- peptides capable of combining specifically with defined protein regions, £•£. , lysozyme.12/ 6. New inhibitors have been developed capable of specific irre- versible or reversible reactions with defined proteins.13/ Some pharmaceutical companies are keeping pace with modern science and technology by attempting to exploit these developments, for instance, in the improvement of inhibitors of dihydrofolate reductase, but most companies are not. Although major pharmaceutical companies usually have the multidisciplinary capability to synthesize and screen new agents, and to take important leads through studies of the mode of action, toxi- cology, clinical pharmacology, and clinical trials, few appear to pos- sess the requisite skills in protein chemistry, or to have incorporated such skills into their programs. On the other hand, although academia is providing most of the basic scientific information, including that on protein structure, academic institutions cannot organize the requi- site multidisciplinary approach necessary to exploit the basic science and technology in this field. It is possible that such integration may be effected through interactions between industry and academia, but it is not yet clear how such integrating efforts might be accomplished. 290

Application of Comparative Biochemistry to Drugs for Developing Countries Many drugs used to treat tropical diseases are highly toxic and have other shortcomings.^/ Nevertheless, we do not wish to minimize possible effects of improvements in delivery systems on drug toxicity. For example, recent experiments have demonstrated that some otherwise highly toxic trivalent arsenicals encased in liposomes can be delivered effectively to leishmaniae-infected liver with relatively little toxi- city ._14_,JLj>/ However, assuming the need for new and better drugs, how shall we consider treatment of tropical parasitic diseases? Parasites causing such diseases fit very well within the scien- tific generalizations presented above: 1. The parasites contain characteristic and specific DNA which determines synthesis of unique essential proteins. 2. They possess essential metabolic systems sensitive to discriminating inhibitors. Some Protozoan Diseases As noted earlier, the dihydrofolate reductase of malarial parasites, such as Plasmodium berghei, differs from that of mammalian and avian hosts in numerous properties.^/ As presented in Table l, these differences permit a substance such as pyrimethamine to exhibit selective toxicity towards the parasite. It would be of great inter- est to determine if inhibition of this enzyme leads to thymineless death in the parasite, as appears to be the case in other organisms.17/ If this is so, the apparent deoxycytidine requirement and purine meta- bolism of Plasmodia 18/ may possibly be manipulated to exacerbate a deficiency for essential thymine. The effects of allopurinol on some Leishmaniae and trypanosomes have pointed to the existence of unique enzymes in these protozoans that can convert the analogue within the parasite to an adenine-1ike nucleotide, as in Figure 2. This newly synthesized analogue can then be inserted into parasite RNA, which leads to inactivation of the para- site._19/ This situation is reminiscent of the effects of inhibitors of herpes virus multiplication activated by virus-induced deoxypyrimidine kinase (Figure l). In the trypanosomiases, it has been known for some time that several African trypanosomes, for example T. brucei and T. rhodesien- sis, depend entirely on glycolysis as their energy source. Unlike anaerobic yeasts, however, the organisms accumulate pyruvate, as shown in Table 2, and must devise new mechanisms to convert NADH to oxidized NAD essential for maintenance of glycolysis. Several such unique 291

FIGURE 2. METABOLIC TRANSFORMATION OF ALLOPURINOL IN T. CRUZI AND MAN* T. cruzi Man Allopurinol ribonucleotide (HPPR-MPf PRPP Allopurinol - Oxipurinol APP ribonucleotide (APPR-MP) APPR-DP APPR-TP RNA PRPP Allopurinol-1 -riboside 4-Aminopyrazolopyrimidine *Human erythrocytes, when incubated with allopurinol, will synthesize trace amounts of l-allopurinol-5•-phosphate. Small amounts of oxipuri- nol-7-riboside are found in human urine. The major metabolic pathway of allopurinol in T. cruzi is to allopurinol ribonucleotide (HPPR-MP)• This is converted by the adenylosuccinate synthetase to the adenylate analogue, APP ribcnucleotide (APPR-MP), which is phosphorylated to the diphosphate (DP) and triphosphate (TP) and incorporated into RNA.JjV 292

CU 4-1 CO 1 1 1 1 1 1 1 1 1 1 4-1 •H CU vO 4J .- H cd c/o •H 1 1 1 1 1 1 1 4- + + [T! O § O 0 3 c/2 GO 0 <; CU Oi 4J 8 CO 4-1 1 + + + + 1 O CO os Oi CU fa cd U § * * * * * * + 1 , , •H 4J C cd J £ 3 M cr 2 i-H i-l CU CO u •H £ CO 1 1 1 1 1 1 +| 1 +| +| C 4-1 M U i I O >J CO 3 £ 3 H l CO S o i H cd 1 1 1 1 1 1 1 1 1 1 „ CO 0 4J •« u W CO <u i-l 0 i- H 4J •H O -J M cu *J 0 u + + + + + + + +1 1 CO od rH cr 0 i-1 ta co 0 S 8 CM O U + |l 1 1 1 1 + + +- +- CO o Q .« O CU as u O- i cd CU M B SB <u Oi 3 co 4-1 tj CU co T3 O • * -H C C Vj CU + 1 COCU-H-HSCU i-l i-l • CU -H CU CO C Cu X O iH CO • •» CN •OJi OC-H-HCOCSON-H CU O09cd33$^Cif99 C W • r^ B Vl ^ C7* CT* iH O ^*4 CU 0 s| V4 bO _^ QJ CU CU ^ O O i— H C! H HHHHHHHHHH 1 293

reactions are known, and these sites are inhibited by antitrypanosomal agents. Thus, ethidium inhibits the unique glycerolphosphate dehydro- genase,^/ and reacts with kinetoplast DNA. The r6--glycerophosphate oxidase and glycerol dehydrogenase of the protozoans also regenerate NAD enzymes. A combination of an Fe chelator, salicylhydroxamic acid, and glycerol prevents this regeneration sufficiently to cause selective damage to the parasites in infected animals.21/ Comparative Biochemistry and Schistosomes Over many years, Dr. Tag Mansour has contributed to our understanding of the many unique biochemical features of helminths. Glycolysis is an essential energy source for survival of the parasite in the mammalian host, although oxygen appears to be required for egg production.^/ As noted earlier, Bueding and J. M. Mansour clarified the special properties of phosphofructokinase 23/ among other enzymes participating in essential glycolytic reactions, but the search for specific inhibitors does not appear to have been viewed as a problem in protein chemistry. It has also been reported that schistosomal eggs contain chitin, a non-mammalian substance.2^/ The presence of eggs in the human urethra and bladder provokes various pathological reactions. Antibiotics which inhibit chitin synthetase and chitinases are known, but experiments with these agents in affecting the course of the disease have not been described. Of course, inhibition of the many reactions related to egg production should be examined. On Leprosy and Mycobacteria We know little of the biological and chemical properties of the etiologic agent causing leprosy. The microorganism is believed to be a very slow-growing mycobacterium, and few workers have successfully cul- tivated the organism in vitro. Cultivation of the organism in a host other than man is an important advance, and its growth in a host even as exotic as the armadillo has facilitated many otherwise impossible studies, particularly serological and immunologic investigations. Despite some advances in these directions, at present we must consider development of chemotherapeutic agents without much solid information about the specific properties of the organism. In this situation we have assumed that the etiologic agent pos- sesses a relatively specific characteristic of Mycobacteria, which is in fact shared with Corynebacteria and Nocardia, jL.j[. , the presence of a cell wall polysaccharide in which about half the sugar residues con- sist of D-arabinose. This sugar is made in bacteria by an enzyme not found in mouse fibroblasts or in yeast ,^5/ and is considered strictly of procaryotic origin. Absence of this enzyme is lethal in E. coli _26/ 294

which contains the essential derivative of D-arabinose, 2-ketodeoxy- octonate. Thus, it can be imagined that unique and inhibitable meta- bolic steps exist in M. Leprae, and in related microorganisms, which are essential for synthesis and insertion of the sugar into the cell wall polysaccharide of the Mycobacterium. Indeed, at least four such unique enzymes should exist, if the synthesis and insertion steps are comparable with the mechanisms known to exist for insertion of sugars into polysaccharides. Conclusions It appears that biochemistry has identified unique metabolic steps essential to survival and multiplication of important parasites. The science and technology applicable to analysis of the origin and struc- ture of essential proteins is ready for a major new direction of attack on the chemotherapy of all infectious disease. The desired methodology is such that, contrary to current approaches by funding agencies, the research budget for tropical disease should be viewed as related also to budgets for research on bacterial and viral disease. The nature of the problem and the approach outlined in this paper require an integrated multidisciplinary attack on parasite-determined proteins, scarcely being attempted at present. A concentration on inhibitors specifically reactive on proteins might well avoid problems of mutagenicity and carcinogenicity of compounds designed to affect nucleic acid metabolism. The time required for such an effort should not be greater than the time usually needed for drug development in a pharmaceutical company. Although the nature of the approach will iden- tify therapeutic leads relatively late in a development process which requires considerable study of the unique essential target protein, search for specificity in the inhibitory process should minimize waste arising from discovery of nonselective toxicity in animal or clinical studies. The funds to be expended might well be much smaller than those spent in futile empirical approaches, as employed in the search for antiviral agents. Finally, a well-organized multidisciplinary effort would provide excellent opportunities to train groups of investi- gators who must ultimately develop in their own countries the drugs to treat the multitude of bacterial, viral, fungal, protozoan, and worm diseases found uniquely in their countries. 295

REFERENCES 1. Janssen PAJ, Thienpont D: In Proceedings of the Conference on Pharmaceuticals for Developing Countries, 1979 2. Bull World Health Org 56:735, 1978 3. Cohen SS: Science 197:43l, 1977 4. Himmelweit F (ed): The Collected Papers of Paul Ehrlich. Chemotherapy. Vol III. New York, Pergamon Press, 1960 5. Albert A: Selective Toxicity. London, Chapman and Hall, ed 5, 1973 6. Mansour TE, Bueding E: Brit J Pharmacol 9:459, 1954 7. Bueding E: Fed Proc 21:1039, 1962 8. North TW, Cohen SS: J Pharmacol Therap, In Press 9. Juan SM, Segura EL, Cazzulo JJ: Int J Biochem 9:395, 1978 10. Hunkapiller MW, Hood LE: Biochemistry 17:2124, 1978 11. Fowler A, Zabin I: J Biol Chem 253:552l, 1978 12. Atassi MZ, Zablock W: J Biol Chem 252:8784, 1977 13. Bey P, Jung M, Metcalf B: Medicinal Chemistry V 115. Amsterdam, Elsevier Scientific Publishing Company, 1977 14. New RRC, Chance ML, Thomas SC, Peters W: Nature 272:55, 1978 15. Alving CR, Steck EA, Chapman WL, Waits VB, Hendricks LD, Swartz GM, Hanson WL: Proc Nat Acad Sci USA 75:2959, 1978 16. Gutteridge WE, Coombs GH: Biochemistry of Parasitic Protozoa. Baltimore, University Park Press, 1977, p 79 17. Cohen SS: Ann NY Acad Sci 186:292, 1971 18. Konigk E: Bull World Health Org 55:249, 1977 19. Marr JJ, Berens RL, Nelson DJ: Science 201:1018, 1978 20. Lambros C, Bacchi CJ, Marcus SL, Hutner SH: Biochem Biophys Res Commun 74:1227, 1977 296

21. Clarkson Jr AB, Brohn FH: Science 194:204, 1976 22. Schiller EL, Bueding E, Turner VM, Fisher J: J Parasitol 61:385, 1975 23. Bueding E, Mansour JM: Brit J Pharmacol & Chemother 12:159, 1957 24. Coles GC: Int J Biochem 4:319, 1973 25. Lim R, Cohen SS: J Biol Chem 241:4304, 1966 26. Rick PD, Osborn MJ: J Biol Chem 252:4895, 1977 297

CELLULAR REGULATORY PROCESSES IN PARASITIC HELMINTHS: SIGNIFICANCE IN DRUG DEVELOPMENT* Tag E. Mansour Cellular regulatory mechanisms of a parasite are essential for maintenance of that parasite's life. For, in addition to adjusting the biochemical machinery of the parasite to meet its needs for energy and reproduction, these control mechanisms must harmonize parasitic bio- chemical and physiological processes with those of the host. Fifteen years ago, I reviewed the pharmacology and biochemistry of parasitic helminths .JL/ The purpose of that review was to emphasize metabolic differences between these organisms and their hosts. Investigations in the field of cellular regulatory biology have since enhanced our know- ledge of many basic principles in enzyme and hormone control among diverse organisms. Progress in understanding control mechanisms in parasites has been meager, because of difficulties in cultivating these parasites in well-defined media, and in the identification and isola- tion of mutants. I shall focus this discussion on three main regula- tory mechanisms: parasite motility, control of metabolism, and chemo- taxis. An important purpose of this article is to draw the attention of persons interested in chemotherapy of parasitic diseases to the action of diverse drugs at various regulatory sites, and possible util- ization of these sites for selection of new antiparasitic agents. Regulation of Motility in Parasitic Helminths A parasitic helminth maintains itself at its specific host site through coordination of its movements. Rhythmical movement of the parasite influences movement of ingested food in the parasite's intesti- nal tract, and its excretory system functions. The important role of motility in maintaining a successful parasitic life has spurred pharma- cologists to use it as a biological parameter for rapid identification of potentially effective new chemotherapeutic agents. In some of these *Parts of this review were supported by Public Health Service Research Grants MH 23464 and HL17976. 298

experiments, the approach taken with isolated organ systems of vertebrates was also used to study parasitic helminth motility. Kymo- graphic recordings of the motility of Ascaris ,2/ and of the liver fluke Fasciola hepatica,3/ were used to study effects of diverse chemi- cal agents on motility. Recently more elaborate systems for monitor- ing motility were devised utilizing instruments mounting multiphoto- cells to measure quantitatively the movement of small parasites such as schistosomes,kj Acetylcholine and serotonin (5-hydroxytryptamine) are two neurotransmitters that have been extensively investigated in parasitic helminths. While studying the effects of drugs on the liver fluke Fasciola hepatica 3/ by means of bioassays, we demonstrated the presence of acetylcholine concentrations in extracts of the worms _5/ which were almost as high as those found in mammalian brains. The presence of both the synthesizing enzyme, acetycholine synthetase, and the hydro- lyzing enzyme, true cholinesterase, was also reported in Fasciola,5/ and in a related trematode parasite, Schistosoma mansoni.6/ Subse- quently, several workers demonstrated the presence of acetylcholine and its two allied enzymes in other trematodes, and in nematodes and cestodes._7/ This evidence, when added to pharmacological data on diverse neuromuscular preparations from these parasites, strongly indicates that acetylcholine or a related compound functions in the neuromuscular activity of these helminths. Acetylcholine in low concentrations has caused Ascaris prepara- tions to contract. However, these preparations were more sensitive to nicotine, which exerted an effect similar to that of acetylcholine. Cholinergic receptors, therefore, do not appear to be muscarinic since the effect of acetylcholine was antagonized by d-tubocurarine but not by atropine.8/ In the liver fluke, the choline esters — carbachol, acetylcholine, and methylcholine — together with physostigmine and prostigmine, relax parasite preparations resulting either in complete paralysis, or in a marked reduction in the amplitude of contraction.^/ The effect of choline esters was reversible. Physostigmine sensitized the preparation to the action of acetylcholine and to carbachol. When added to the fact that acetylcholine, true cholinesterase, and ace- tylcholine synthetase are present in these organisms, these results strongly indicated the presence of cholinergic receptors in these trematodes. llhese receptors are peripherally located and appear responsible for relaxing or paralyzing the motility of these organisms. The evidence indicates that acetylcholine receptors in trematodes may differ from those found in nicotinic or muscarinic mammalian synapses.J) This emphasizes an important conclusion; namely, that regulatory pro- cesses at the cholinergic receptors of these parasites appear to differ from those of the host. In 1949, we discovered that rhythmical activity of both intact and degangliated preparations from liver flukes were stimulated by sym- pathomimetic amines, particularly those related to amphetamine.^/ We 299

subsequently reported that rhythmical movement of the liver fluke was stimulated by serotonin, and by lysergic acid diethylamide (LSD) and related indoleamines (9,10). The effect was peripheral, and was not mediated through the parasite's central ganglion. Bromolysergic acid diethylamide, an analog of LSD with a bromine atom at the 2" position, depressed rhythmic movement and antagonized the stimulant action of serotonin and LSD. The results suggested the presence of serotonin receptors in these organisms. Pharmacological evidence that serotonin is a putative excitatory neurotransmitter in the liver fluke j^,_10/ raised the question of whether a similar role can be demonstrated in other flat worms. This was in fact indicated by the finding that serotonin stimulated rhythmic movement of the following species: Schistosoma mansoni, Chlonorchis sinesis, and Taenia pisiformis.l/ These findings were further verified in Schistosoma mansoni by two other investigator groups ,_L1,_12/ and also apply to the cestode, Mesocestoides corti.13/ Serotonin is an ubiquitous indoleamine in invertebrate tissues. J-4^15/ Erspamer and Welch _14_i_15/ demonstrated its presence in several invertebrate nervous systems. From our laboratory, and based on results from a bioassay procedure _l,_16_/ and fluorometric methods ,17/ we reported that the indoleamine is present in the liver fluke, Fasciola hepatica.l ,16,17/ Andreini et a_l._18/ published data suggesting that the indoleamine in the fluke differs spectrophotofluorometrically from serotonin. The failure of Chou et a^.lj)/ to demonstrate indoleamine in the liver fluke may be due to their use of unstarved parasites which contain large amounts of caecal contents that include lysine. The same group of investigators noted that this amino acide gives a spectro- fluorometic signal similar to that of serotonin.2Q_ f Intact liver flukes were capable of synthesizing the indoleamine from 5-hydroxy- tryptophan but not from tryptophan._l,_16_,_17/ The presence of 5-hydroxy- tryptophan decarboxylase and its activation by pyridoxal phosphate was further demonstrated in cell-free extracts from the flukes _l,_16/ using bioassay procedures. The presence of serotonin and of an uptake mech- anism was reported in schistosomes.^1/ The data implicate serotonin or a related indolamine as a putative neurotransmitter in the regula- tion of neuromuscular activity in parasitic flat worms. Cheniotherapeutic Agents that Affect Parasite Motility Several studies indicate that particular chemotherapeutic agents are effective because they modify one or another regulatory mechanism involved in motility. Santonin, an obsolete anthelmintic once used exclusively against nematodes, was found by Baldwin 2_/ to paralyze only the anterior neuromuscular preparations in Ascaris at low concen- trations, and to stimulate intermediate preparations. Santonin may, therefore, deprive the worm of coordinating impulses which come from the central nerve ring. This action is presumed to result in the 300

parasite's not being able to maintain its normal position in the host's intestine, and therefore becoming subject to expulsion. More recent studies on the mode of action of piperazine on Ascaris add evidence to the idea that effective chemotherapeutic agents may act upon regulatory mechanisms of neuromuscular function. Standen 22/ showed that Ascaris were paralyzed when incubated with piperazine. Subsequently, Norton and DeBeer 23/ showed that acetylcholine-induced contractions in Ascaris preparations could be blocked by piperazine and by d-tubocurarine. Piperazine may cause paralysis of Ascaris by blocking at the neuromus- cular junction. The anthelmintic agent may be a pharmacological analog of a natural inhibitory neurotransmitter. The molecular identification of such a natural transmitter is an important problem that deserves further investigation. Methyridin (2-B-methoxyethylpyridine), an anthelmintic with high activity against nematodes, was reported by Broome 24/ to act on the neuromuscular junction of Ascaris. According to this finding, methy- ridin and acetylcholine cause rapid paralysis of rehythmic movement. The effect of both compunds was reversed by d-tubocurarine and by piperazine. Methyridin may therefore produce a depolarizing effect on nematode muscle closely resembling that produced by acetylcholine. A similar effect was demonstrated in mammalian tissues, but only at very high drug concentrations. The selective effect of this drug as a chemo- therapeutic agent seems to depend on a differential sensitivity of nematode neuromuscular receptors to its action. Hycanthone is a new and highly schistosomicidal agent. Chou et al. 25/ reported evidence implicating serotonin receptors in the chemo- therapeutic effect of this agent. Hycanthone treatment of the host results in a loss of the ability of the parasite to store serotonin in neuronal structures and, consequently, in its ability to coordinate its motility. Accordingly, hycanthone may interfere with a regulatory mechanism that is essential for maintenance of the parasite in its normal habitat. Among anthelmintics effective against cestodes, arecoline may be used as a classical example. This muscarinic cholinergic agent, al- though not currently in clinical use, was the drug of choice a quarter of a century ago. Because of its muscarinic effect on the host, it was thought that the drug exerted its anthelminitc action as a result of its purgative effect. Batham 26/ showed that this was not the case since, when the drug was injected subcutaneously, it caused purgation but not worm removal. Using neuromuscular preparations of Taenia, Batham showed that it causes relaxation and eventual paralysis of the worm at low concentrations. In vivo experimentation supports the view that some anthelmintic drugs owe their effectiveness to a primary action on motility. In 1944, Rogers 2T/ showed how tetrachloroethylene, a drug that was extensively used against Ancylostoma, caused the nematode parasite 301

Nippostrongylus muris to leave the mucus of the rat's intestinal wall, after which they were expelled while still alive by movement of gut contents. A closely related phenomenon was observed by Standen 28/ in his early studies on antischistosomal agents. This is referred to as "the parasite shift effect." It can be demonstrated to follow treatment of infected mice with an antischistosomal agent. A "shift" can be observed in the distribution of the parasites from the mesente- ric vessels to the liver. Standen explained this as due to interfer- ence by these agents with the muscle tone of the parasites. The end result is that they are unable to hold on to the walls of the blood vessels, and are gradually swept into the liver with the portal blood. Death of the schistosomes then occurs in the liver, according to Standen's histological studies.28/ Serotonin Regulation of Metabolism in Flukes Although information on hormonal control of metabolism of parasites is scanty, the evidence indicates that control of metabolism in para- sites may be mediated through hormones that differ from those of the host. The involvement of cyclic 3',5'-AMP as a second messenger has already been reported in some parasites. The liver fluke Fasciola hepatica is being used in our laboratory as a model to study regulation of carbohydrate metabolism in trema- todes. Fasciola is an anaerobic organism that metabolizes glucose at a high rate, and converts it almost quantitatively to volatile fatty acides and CO with formation of only small amounts of lactic acid.29/ Incubation of the organism with serotonin causes a marked increase in lactic acid production.3^/ In addition to its stimulatory effect on glycolysis, serotonin increases glycogen breakdown when glucose is omitted from the fluke medium, and activates glycogen phosphorylase and adenylate cyclase .^30,317 None of these effects can be mediated by epinephrine or norepinephrine. The effect of serotonin on glycolysis does not seem restricted to Fasciola. The indoleamine, and its agonists methylsergide and dihydroergotamine, increases lactic acid production by schistosomes.32/ Furthermore, Higashi et al._3_3/ reported that adenylate cyclase from schistosomes is activated by serotonin. Thus, serotonin affects carbo- hydrate metabolism and adenylate cyclase in the liver fluke and in schistosomes similarly to epinephrine in some mammalian tissues. This is a very good illustration of a difference between the parasite and the host that can be utilized for planning the search for a chemothera- peutic agent. Serotonin and Cyclic AMP in the Liver Fluke Adenylate cyclase in the liver fluke 31,34-36/ is a membranous 302

enzyme with multiple components for its control. While basal activity (without activators) is low, it is markedly activated in the presence of serotonin. Activation by serotonin 35/ is dependent on the presence of GTP.^35/ Such a requirement is also seen with activation of adenylate cyclase by mammalian hormones. The specificity of serotonin activation of the enzyme was recently confirmed from studies on structure-activity relationships with diverse analogs ._35/ Serotonin appears to activate adenylate cyclase. Although LSD and its derivatives are poor agonists compared to the indoleamines, they demonstrate greater affinity to serotonin sites .^IV D-LSD was the most potent derivative. Any substi- tution other than of the diethyl group on the amide nitrogen decreased the potency of the derivative. Either D- or L-LSD act as antagonists of serotonin activation with very low affinity for L-LSD. Thus, both activation of adenylate cyclase and antagonism of 5-HT are stereo-speci- fic processes. BOL, which is not an agonist to adenylate cyclase acti- vation, competes for the 5-HT activation and inhibits basal (non-acti- vated) enzyme activity. BOL was the most potent antagonist with a KJL value in the nanomolar range.35 Cyclic 3',5'-AMP phosphodiesterase is the only enzyme known to hydrolize cyclic AMP to 5-AMP. Both adenylate cyclase and phosphodi- esterase regulate the concentration of cyclic AMP in the cell. Phos- phodiesterase was found in fluke homogenates ^3_7/ and appears to be involved in regulation of cyclic AMP concentrations in this organism. Phosphodiesterase inhibitors increase motility like serotonin and its analogs._3_7/ The molecular basis for this effect does not appear to be simple inhibition of the phosphodiesterase, thus raising the concentra- tion of cyclic AMP. Only one of these inhibitors, isobutyl methyl- xanthine, increaed the concentrations of cyclic AMP in the flukes. Indeed, many of these agents antagonized serotonin-mediated rises in concentrations of endogenous cyclic AMP. While direct participation of cyclic AMP in fluke motility has not been demonstrated, phosphodi- esterase appears to offer a site for chemical agents to influence motil- ity of the parasite. Some effects of hormones are known to be mediated by cAMP through activation of protein kinases which phosphorylate specific proteins. Epinephrine activation of glycogen phosphorylase offers a classic example of such a control mechanism. Recent studies in our laboratory by Gentleman «^t £l._38/ demonstrated the presence of a protein kinase both in the particulate and in the supernatant fractions of Fasciola. The fluke protein kinase was highly sensitive to cAMP. Incubation of flukes with serotonin activated the protein kinase. Under these con- ditions, the enzyme was highly active without cAMP. Interestingly, and under conditions which favor an antagonistic effect on the adenylate cyclase system in the flukes, LSD also antagonized activation of the protein kinase._38/ The results may indicate that, as in mammalian cells, some physiological effects of serotonin are mediated by cyclic AMP activation of protein kinase. Such an effect may also apply to other trematodes, since S. mansoni was reported to contain this enzyme 303

together with an adenylate cyclase activated by serotonin.^3/ Sub- strates for the protein kinase have still to be isolated and identified. Pharmacologic Manipulation of Phosphofructokinase Studies have been carried out on enzymes catalyzing rate-1imiting reactions to increase understanding of the details of their regulation. Usually, these enzymes are allosteric proteins and exhibit complex kinetics complementary to the metabolic needs of the cell. Phospho- fructokinase is one enzyme that has attracted our attention. This enzyme catalyzes the physiologically irreversible phosphorylation of fructose-6-P to fructose-1,6-P2 with ATP as the phosphate donor. The enzyme was shown to regulate glycolytic flux in higher organisms (for references see 39). Evidence was presented indicating that serotonin activates phosphofructokinase in the liver fluke as it activates adenylate cyclase.4£,^1,^2/ Control of the enzyme appears to involve conversion of a low molecular weight inactive form of phosphofructo- kinase to a high molecular weight active form. The fluke enzyme, once activated, is subject to all the allosteric kinetics relevant to mam- malian phosphof ructokinase._3£,4_3/ This includes inhibition by one of its substrates, ATP. The ATP-inhibited enzyme can be "de-inhibited" by cyclic AMP and fructose-6-P. Kinetic studies with the active enzyme showed that each of the three ligands (fructose-6-P, ATP, and cyclic AMP) could alter the saturation curve of the others. These kinetics were considered in terms of the Monod-Changeux-Wyman model for allo- steric proteins .^4. Thus, the parasite enzyme is well endowed with all the control mechanisms necessary for reversible conversion from active to inactive form through changes in the tertiary and in the quaternary structure of the enzyme, a classic example of a regulatory enzyme. Because of its critical role in regulation of glycolysis, phospho- fructokinase has been implicated as the site of action of an important group of antischistosomal agents, the antimonials. Utilizing a cell free extract system for measuring glycolysis, we observed that phospho- fructokinase is a rate-1imiting enzyme in schistosomes,45^4jj/ and that its inhibition can account for the effect of these antischistosomal agents. The parasite enzyme is comparatively more sensitive to inhibi- tion by these antimonials than is the mammalian phosphofructokinase.A5/ Similarly, selective inhibition by antimonials on adult filariids was recently reported. Dipetalonema witei (=viteae) and Brugia pahangi were reported to be metabolically similar to the schistosomes in that they are homolactate fermenters.^7,^8/ Helminths, such as Ascaris and Hymenolepis diminuta, which are not homolactate fermenters also exhib- ited phosphofructokinase activities which were susceptible to inhibi- tion by stibophen. This is in direct contrast to phosphofructokinases derived from mammalian tissues which are considerably more refractory to inhibition by antimonials._48/ These findings, when added to our results with schistosomes ,^5 suggest a broad spectrum of activity by 304

antimonials against parasitic helminth phosphofructokinases. This enzyme model is amenable to selective inhibition by chemotherapeutic agents. Other Regulatory Sites Affected by Anthelmintics Schistosomes seem to depend primarily, if not exclusively, on preformed purines to satisfy their needs for purine nucleotides.49/ Utilizing this finding, Jaffe 50/ found that the purine analog, tuber- cidin (7-deazoadenosine), exerted a potent antischistosomal effect due to inhibition of the utilization of adenosine for adenine nucleotide formation.51/ One enzyme system that appears essential for many trematodes, nematodes and cestodes that ferment glucose to succinate or to volatile fatty acids is fumarate reductase, a component of the succinic dehydro- genase complex. Since these parasites are predominantly anaerobes, the enzyme plays an important role in the regeneration of NAD ._52-_55_/ Inhi- bition of this enzyme markedly suppresses the viability of these para- sites. _56-59/_ Metzger and Duwel 60/ recently demonstrated that the mechanism of action of 2,6-dihydroxy-3,5-dichlorobenz-4'-chloroanilide (DDBA) as a flukicide may be ascribed to a primary inhibition of fumarate reductase within the succinic dehydrogenase complex in the parasite.57/ Studies on the biochemistry of egg formation should be of importance in developing drugs to interfere with parasite egg laying capacity. Bennett and Gianutsos 61/ reported that disulfuran raises the levels of dopamine when given to schistosome-infected mice and then reduces the levels of norepinephrine to almost undetectable concentra- tions in schistosomes. Although changes in motility of the parasites were not observed, such treatment resulted in abnormal egg production. This effect appears related to inhibition of the enzyme phenol oxidase. This enzyme was first demonstrated in trematodes in Fasciola hepatica, 62/ and was subsequently found in schistosomes.(yY Phenol oxidase has a variety of substrates, and may be involved in the oxidation of an as yet undetermined catechol substrate necessary for egg formation. The effect of disulfuran on egg formation by the parasites may be due to inhibition of the activity of this enzyme. Studies on the biochemistry of egg formation in the parasite may elucidate other sites amenable to selective inhibition by chemical agents. Chemotaxis Chemotaxis is one of the most primitive processes enabling an organism to respond to favorable and noxious agents in the environment. The process is being extensively investigated in prokaryotes ,^3/ and appears to involve a receptor-mediated stimulus event, a sensing 305

mechanism that is temporal in nature, and movement of the organism in response to the stimulus. Studies on eukaryotic cells have not ad- vanced as far as those in bacteria. Chemotaxis of leukocytes to vari- ous complement-derived factors and byproducts of the inflammatory response ,j>4,j^5 and to n-formyl methionine-containing peptides has recently been reported. In addition, the process of aggregation in the cellular slime mold, Dictyostelium discoideum, has been shown to be mediated via chemotactic responses to oscillating gradients of cyclic AMP produced by the starving amebae.66,67/ We have recently studies chemotaxis, using the acellular slime mold, Physarum polycephalum.68,69/ The organism provides both an objective and quantitative model for the measurement of chemotaxis, and information on this organism may provide additional insight into the nature of this important process. Chemotaxis in parasitic helminths appears intricately involved in completion of the life cycle. Migration of trematode miracidia to intermediary snails, and attraction of the motile cercariae to mammalian hosts of some trematodes, are classical examples of chemotaxis in the life of these parasites. A study has been initiated on the details of how miracidia and cercariae find their hosts. Recent work reported by Maclnnis 70/ identified chemicals from mammalian skin which initiated penetration responses by Schistosoma mansoni cercariae. Earlier studies by the same group implicated amino acids and other compounds as attractants of S. mansoni miracidia to the snail host.^71./ Since eigh- teen amino acids have been reported to be excreted by the snail, it remains to be seen whether a few amino acids can be identified as the real attractants in this large mixture. The value of information gained from these studies for development of drugs and other methods for controlling schistosomiasis and other trematode infections 71/ is obvious. Analogs of these amino acids may act as antagonists instead of agonists, and may result in disrupting an important system for completion of the life cycle. Misleading miracidia into migrating to a non-favorable environment may be sufficient to terminate their life cycle. Research on biology of chemotaxis has been extended to studies on snail vectors of parasites. Recently, Maclnnis' group has investigated chemicals emanating from food sources which can attract or trap the snail._7_2/ Fractionation of lettuce by these workers revealed that the portion containing free amino acids was the primary attractant for Biomphalaria glabrata, the snail host for Schistosoma mansoni. These amino acids have been identified as glutamate and proline. A trapping mechanism, following attraction of the snails, may be devised by provid- ing these attractants as baits. Similarly, molluscicides may be effec- tively used if they are released in the water following a prolonged release of attractants such as proline and glutamate. 306

The above overview illustrates that, despite spectacular advances in knowledge concerning chemotaxis in prokaryotes during the last decade, information available on chemotaxis in parasites is very limited. Intricate details of this process may expose sites amenable to pharmacologic manipulation. The available information concerning cellular regulatory processes in parasitic helminths is scanty, and much research is needed in this important area. Workers in this field may benefit significantly from the recent explosion of knowledge of cellular regulation in bacterial and mammalian cells. Studies on cellular regulatory processes dis- cussed above, and involving motility and metabolism, support the view that the the nature of these regulatory processes in the parasite may be different from that of the host. An important aspect of this thesis is that these processes may be exploited toward a strategy for chemo- therapy of the infections caused by these parasites. 307

REFERENCES 1. Mansour TE: Advances Pharmacol 3:129-165, 1964 2. Baldwin E: Parasitology 35:89-11l, 1943 3. Chance MRA, Mansour TE: Br J Pharmacol 4:7-13, 1949 4. Hillman GR, Senft AW: J Pharmacol Exp Ther 185:177-184, 1973 5. Chance MRA, Mansour TE: Br J Pharmacol 8:134-138, 1953 6. Bueding E: Br J Pharmacol 7:563-566, 1962 7. Von Brand T: Biochemistry of Parasites. New York, Academic Press, 1973 8. del Castillo J, de Mello WC, Morales T: Br J Pharmacol 22:463-477, 1964 9. Mansour TE: Br J Pharmacol 12:406-409, 1957 10. Beernink KD, Nelson SD, Mansour TE, Int J Neuropharmacol 2:105-112, 1963 11. Barker LR, Bueding E, Timms AR: Br J Pharmacol Chemother 26:656-665, 1966 12. Hillman GR, Senft AW: J Pharmacol Exp Ther 185:177-184, 1973 13. Hariri MJ: J Parasitol 60:737-743, 1974 14. Erspamer V: Pharmacol Rev 6:425-487, 1954 15. Welsh JH, Moorhead MJ: J Neurochem 6:146-169, 1960 16. Mansour TE, Lago AD, Hawkins JL: Fed Proc 16:319, 1957 17. Mansour TE, Stone DB: Biochem Pharmacol 19:1137-1146, 1957 18. Andrieni GC, Beretta C, Faustini R, Gallina G: Experientia 26:166-167, 1970 19. Chou T-CT, Bennett J, Bueding E: J Parasitol 58:1098-1011, 1972 20. Tomosky-Sykes TK, Jardine I, Mueller JF, Bueding E: Anal Biochem 83: 99-108, 1977 308

21. Bennett JL, Bueding E: Mol Pharmacol 9:31l-319, 1973 22. Standen OD: Brit Med J 2:20-22, 1955 23. Norton S, de Beer EJ: Am J Trop Med Hyg 6:898-905, 1957 24. Broome AWJ: Drug Parasites and Hosts. Edited by LG Goodwin, RH Nimmo-Smith. Boston, Massachusetts, Little, Brown, 1962, p 43 25. Chou T-CT, Bennett JL, Pert C, Bueding EJ: Pharmacol Exp Ther 186:408-415, 1973 26. Batham EJ: Parasitology 37:185-191, 1946 27. Rogers WP: Parasitology 36:98-109, 1944 28. Standen OD: Ann Trop Med Parasitol 44:26-43, 1953 29. Mansour TE: Biochein Biophys Acta 34:456-464, 1959 30. Mansour TE: J Pharmacol Exp Ter 126:212, 1959 31. Mansour TE, Sutherland EW, Rall TW, Bueding E: J Biol Chem 235:466-470, 1960 32. Hillman GR, Olsen NJ, Senft AW: J Pharmacol Exp Ther 188:539-535, 1974 33. Higashi GI, Kreiner PW, Keirns JJ, Bitensky MW: Life Sce 13:121l-1220, 1973 34. Abrahams SL, Northup JK, Mansour TE: Mol Pharmacol 14:804-819, 1978 35. Northup JK, Mansour TE: Mol Pharmacol 14:804-819, 1978 36. Northup JK, Mansour TE: Mol Pharmacol 14:820-833, 1978 37. Mansour TE, Mansour JM: Biochem Pharmacol 26:2325-2330, 1977 38. Gentleman S, Abrahams SL, Mansour TE: Mol Pharmacol 12:59-68, 1976 39. Mansour TE: Curr Top Cell Regul 5:l-46, 1972 309

40. Mansour TE: J Pharmacol Exp Ther 135:94-10l, 1962 41. Mansour TE, Mansour JM: J Biol Chem 237:629-634, 1962 42. Stone DB, Mansour TE: Mol Pharmacol 3:16l-176, 1967 43. Stone DB, Mansour TE: Mol Pharmacol 3:177-187, 1967 44. Monad J, Wyman J, Changeux J-P: Mol Biol 12:88-118, 1965 45. Mansour TE, Bueding E: Br J Pharmacol 9:459-462, 1954 46. Bueding E, Mansour JM: Bri J Pharmacol 12:159, 1957 47. Wang, EJ, Saz HJ: J Parasitol 60:316-32l, 1974 48. Saz HJ, Dunbar GA: J Parasitol 61:794-80l, 1975 49. Senft AW: J Parasitol 56:314, 1970 50. Jaffe JJ, Meymarian E, Doremus HM: Nature 230:408-409, 1971 51. Ross AF, Jaffe JJ: Biochem Pharmacol 21:3059-3069, 1972 52. Bryant C: Adv Parsasit 8:139-172, 1970 53. Metzger H: Z Parasitkde 34:27l-295, 1970 54. Prichard RK, Schofield P: Comp biochem Physiol 25:1005-1019, 1968 55. De Zoeten LW, Posthuma D, Tipker J: Hoppe-Seyler's Z Physiol Chem 350:683-690, 1969 56. van den Bossche H, Janssen PAK: Life Sci 6:178l-1792, 1967 57. van den Bossche H, Janssen PAJ: Biochem Pharmac 18:35-42, 1969. 58. Schiebel LW, Saz HJ, Bueding E: J Biol Chem, 243:2229-2235, 1968. 59. Prichard RK: Nature 228:684, 1970 60. Metzger J, Duwel D: Int J Biochem 4:113-143, 1973 61. Bennett JL, Gianutsos G: Biochem Pharmacol 27:817-820, 1978 310

62. Mansour TE: Biochem Biophys 30:492-500, 1958 63. Adler J: Ann Rev Biochem 44:43l-456, 1975 64. Boyden S: J Exp Med 115:453-466, 1962 65. Schiffman E, Corcoran BA, Wahl SM: Proc Nat Acad Sci USA 72:1059-1062, 1975 66. Konijn TM, van de Meene JGC, Bonner JT, Barkley DS: Proc Nat Acad Sce USA 58:1152-1154, 1967 67. Gerisch G, Wick U: Biochem Biophys Res Communi 65:364-370, 1975 68. Kincaid RL, Mansour TE: Exp Cell Res 116:365-375, 1978 69. Kincaid RL, Mansour TE: Exp Cell Res 116:377-385, 1978 70. Maclnnis AJ: Nature 224:122l-1222, 1969 71. Maclnnis AJ, Bethel WM, Cornford EM: Nature 248:36l-363, 1974 72. Uhazy LS, Tanaka RD, Maclnnis AJ: Science 201:924-926, 1978 311

RATIONAL DEVELOPMENT OF NEW DRUGS FOR TRYPANOSOMIASIS Anthony Cerami and Steven R. Meshnick Although development of therapeutic agents for parasitic diseases has been largely neglected in the past fifty years, this has not always been the case. In fact, as is evident from the classical studies of Ehrlich on malaria and trypanosomes, these organisms served as the model for development of chemotherapeutic agents in general.jY In recent years, however, the pharmaceutical industry lacked incentive to find new drugs for diseases of the developing world. This has been brought about by two factors: the substantial amount of capital neces- sary to evaluate safety and efficacy of possible new agents; and the limited resources of the developing world. The new nationalism and awareness of problems of the developing world in the last few years has made it apparent that additional efforts must be undertaken to alleviate this problem. The means for overcoming this impasse, however, are not obvious. In the past few years, both WHO and NIH have launched major programs to study parasitic diseases. But it is doubtful whether university-based research pro- grams can, by themselves, achieve the desired ends, since universities often lack facilities for evaluation of drug safety and experience in this subject. Thus, there is a definite need for participation by the pharmaceutical industry in development and eventual distribution of new drugs. It therefore becomes imperative to develop a holistic approach which would facilitate application of modern chemistry and biology to understanding parasitic diseases and developing means for their con- trol. One possible solution would be cooperative ventures consisting of governmental, pharmaceutical, and academic groups. This is particu- larly important since it is doubtful that these new drugs would ever be as important a source of revenue as a new broad spectrum antibiotic. One of the ways that cost of drug development could be signifi- cantly reduced is by applying modern chemistry and biochemistry to rational design of new drugs for parasitic diseases. Rational design of drugs for treatment of orphan diseases is in fact an area of great interest to our laboratory._2,^3/ We have recently tried to apply these 312

principles to elucidation of new drugs for parasitic organisms and have studied African trypanosomes (Trypanosoma brucei and Trypanosoma congo- lense) as model organisms«^,^5,^>/ Our approach was to elucidate a bio- chemical difference between trypanosomes and their host, and to design drugs which could take advantage of this difference. The biochemical difference we have exploited is inability of trypanosomes to synthesize heme._/y As a result of this deficiency, and the avid binding of heme to serum proteins in mammalian hosts, the bloodstream form of African trypanosomes has no detectable heme or hemoproteins such as cytochrome or catalase. This in turn leads to an accumulation of intracellular hydrogen peroxide in these organisms which renders them susceptible to killing by agents that promote homolytic cleavage of hydrogen peroxide to yield hydroxy (OH.) or hydroperoxy (HOO.) radicals. These radicals can react with unsaturated lipids and other cell constituents, thereby leading to cell destruction.4/ Intracellular hydrogen peroxide arises in an adventitious manner from the x-glycerol phosphate oxidase complex in the rudimentary mito- chondria of the trypanosome. Approximately l-3 percent of the oxygen consumed by these parasites appears as hydrogen peroxide which leads to a final concentration of approximately 70jjM. In addition to lacking catalase these organisms also lack other enzymatic means for destroying this waste product, and depend upon diffusion of this potentially lethal metabolite from the cell to its surroundings where it can be dis- posed of by mammalian enzymes. Figure 1 shows that portion of the bio- chemistry of the parasite in which production of hydrogen peroxide is thought to arise, and depicts those sites where we have sought to inter- vene to increase the concentration of intracellular hydrogen peroxide or its homolytic cleavage products._5/ We have found that hematoporphyrin D and meso-tetra 4-sulfonato- phenyl porphine (TPPS4) are free radical initiators, and cure trypano- some-infected mice. It is of interest that neither of these compounds was active in vitro. Analysis of this phenomenon revealed that inser- tion of zinc was necessary for cleavage of hydrogen peroxide to occur with subsequent killing of the organisms. This conversion was cata- lyzed by the zinc complex, probably via removal of a "77- electron from the porphyrin ring. This is unlike the reaction of iron porphyrins with H 0 where oxidation occurs at the metal atom. Knowledge of the mechanism of action of the porphyrin analogues permitted further intervention._5/ As seen in Figure l, methods for increasing hydrogen peroxide should make porphyrin derivatives more effective. It has been possible to increase the rates of both oxygen consumption and hydrogen peroxide production in trypanosomes in vitro by addition of naphthoquinones. Presumably, naphthoquinones act as coenzyme Q analogues and cycle as shown in Figure 2. The naphthoqui- nones are reduced to the corresponding quinols by the dehydrogenase, and then nonenzymatically reduce oxygen to form superoxide and hydrogen peroxide. Meanwhile, the quinone is reformed and can be cycled again. 313

FIGURE 1. Glycerol phosphate oxidase NAD odGP Glycolysis JU5PDH) NADH DHAP- I Dehydro- CoQH2? CoQ? Oxidase 1 H,0 Trapped Damage to by GSH cell constituents The addition of sublytic concentrations of both napthoquinones and heme leads to a synergistic lysis of the organisms iri vitro. Although naphthoquinones are trypanocidal in vitro both alone and in combination with porphyrin analogues, we have not yet been able to identify a naph- thoquinone active in vitro either by itself or in combination with various porphyrins. We are continuing a search for potentially active naphthoquinones. Another method of rendering trypanosomes more susceptible to radicals generated by porphyrin-induced decomposition of hydrogen perox- ide would be to decrease the level of intracellular radical scavengers 314

FIGURE 2. Dehydrogenose OH H202 in the trypanosome. Reduced glutathione is believed to play such a protective role in these organisms. Incubation of isolated blood- stream forms of T\ brucei with the trivalent arsenical, melarsenoxide, led to a decreased intracellular concentration of glutathione which imparts to the organisms an increased susceptibility to heme lysis. In fact we have noted synergism between arsenicals and hematoporphyrin D in vitro. Since both hematoporphyrin and Mel B can cross the blood/ brain barrier, an evaluation of the combination of both agents in ad- vanced forms of sleeping sickness is warranted. From our in vitro studies to date, it is apparent that the combination of a series of drugs which can increase hydrogen peroxide 315

production, cause homolytic breakdown of hydrogen peroxide, and decrease glutathione concentration should be an effective means of selectively killing these organisms. As noted earlier we began our studies with African trypanosomes as model organisms in which to look for biochemical differences between host and parasite. In addition to the African trypanosome, it appears that Trypanosoma cruzi, which causes Chagas' disease in approximately seven million people in South America, also lacks the ability to enzymatically destroy hydrogen peroxide.8/ Furthermore, naphthoquinones have been found which cause increased production of this potentially lethal byproduct in vitro.9/ This, coupled with the fact that hematophorphryin D and TPPS are active against Leishmania donovani,10/ emphasizes the similarity of the trypanosomatids which cause human and animal diseases. New agents, which take advantage of the biochemical differences outlined above, may be useful in several of these diseases. Further studies, like the one described above, may lead to identi- fication of other biochemical differences which could be exploited to develop new drugs. This kind of investigative process is especially well carried out by academic scientists who are intrigued by basic bio- chemical studies and who are funded by agencies such as NIH and WHO. Once a biochemical difference is identified, it could perhaps be handed over to organic chemists in the pharmaceutical industry who could best synthesize and identify more efficacious anlogues. Because much of the cost of drug development today is consumed in safety evaluation of the new compound, sharing of these costs between the pharmaceutical indus- try and public health organizations might achieve this last and impor- tant objective. In the next fifty years the number of parasitic dis- eases controlled through chemotherapy can be significantly higher than it is today. 316

REFERENCES 1. Hawking F: Experimental Chemothapy. Edited by RJ Schnitzer, F Hawking. New York, Academic Press, 1963, I, pp l-24 2. Gillette PN, Peterson CM, Lu YS, Cerami A: New Engl J Med 290: 654-660, 1974 3. Grady RW, Graziano JH, Whitel GP, Jacobs A, Cerami A: J Pharmacol & Exper Therap 205:757-765, 1978 4. Meshnick SR, Chang KP, Cerami A: Biochem Pharmacol 26:1923-1928, 1977 5. Meshnick SR, Blobstein SH, Grady RW, Cerami A: J Exp Med 148:569-579, 1978 6. Meshnick SR, Grady RW, Blobstein SH, Cerami A: J Pharmacol Exp Therap 207:104l-1050, 1978 7. Chang KP, Chang CS, Sassa S: Proc Natl Acad Sci USA 72:2979-2985, 1975 8. Docampo R, de Boiso JF, Boveris A, Stoppani AOM: Experientia 32:972, 1976 9. Boveris A, Docampo R, Turrens JF, Stoppani AOM: Biochem J 175:l-9, 1978 10. Chance M, Meshnick SR, Cerami A: (unpublished observations). 317

IMMUNOLOGY AND PARASITIC DISEASES* John R. David In considering some infectious diseases in the developing world that are the subject of this meeting, it became clear that, as in the course of political events, one must know the past to understand the future. And the history of these diseases tells us that we must now make a multifaceted attack if we hope to bring them under control. This is best illustrated by the history of attempts to control malaria, so beautifully set forth in a recent book by Gordon Harrison.jY For many years immunologic study of malaria was in the doldrums, because it seemed so obvious that insecticides and drugs were the agents which would control this disease. Now that mosquitos are thriving on DDT, and drug resistant strains of malaria are occurring, development of a vaccine seems more reasonable, and work to develop one is underway in several laboratories. Indeed, at a time when one of the scourges of mankind, smallpox, essentially has been eradicated from the world by a vaccine, it is small wonder that one turns to immunology for some help in combatting other infectious diseases afflicting the developing world. Although three commercial vaccines are available against parasites of animals, cattle and sheep lung worm, and dog hookworm, it is dis- appointing that there are none for protozoa or helminths infecting man. But vaccines against parasites are not easy to come by, because these parasites have evolved very effective ways of living in their hosts, and of evading their hosts' immune defenses. Parasites have evolved at least four methods of evading immune attack. Some, like schistosomes, incorporate host antigens onto their surface and masquerade as the host.J^JJ/ Some, like the African trypano- somes, can change their surface antigens when attacked by antibody, an an amazing biological adaptation. 4^ J3/ Some turn on a subset of T lymphocytes, whose function it is to suppress the immune response.])/ And, finally, there are suggestions that some parasites may develop *Supported in part by The Rockefeller Foundation, The Edna McConnell Clark Foundation, and The Wellcome Trust 318

innate resistance.]_,^l In other words, they are still recognized as foreign by the host, but have developed changes on their surface which make them impregnable to attack by the host's humoral and cellular defense mechanisms. At this point, I should define an important concept relevant to this subject, that of concomitant immunity. When the host is first invaded by a parasite, it has no specific immunity. The parasi__ induces an immune response, but it also manias to evade effe; . • . * this response by one of the four mechanisms described above and lives relatively unperturbed in the host. If another parasite of the same species subsequently invades the host, the second invader will be destroyed by the immunity induced by the first infection. In this way, the balance required for parasitism is preserved, the parasite lives on, and the host is not overwhelmed. I would like now to consider the short-term and long-term goals of immunologic research as it relates to our current problems. There are at least three short-term goals. We must obtain more information on the role and mechanism of immunity in specific parasite diseases, develop better serodiagnostic tools, and obtain more basic information on regulation of the immune response in general. The latter goal I will return to at the end of this presentation. Work in this area is essential if we are to have a rational basis for developing effective vaccines. Why, you may ask, is there a need for more basic studies when Jenner knew nothing about T or B lymphocytes, let alone macrophages, eosinophils or antibodies? The answer is simple. First, immunity to smallpox, and to many other viruses, is obvious, direct, and long- lasting. Second, the attenuated virus can be grown easily. Neither of these situations pertains in the case of parasites. If we are to develop vaccines and useful immunodiagnostic tools, and attempt to deal with the tissue damage resulting from the immune response, we must be able to answer certain types of questions. Is there immunity to T\ cruzi in Chagas' disease? In that disease, is the first infection responsible for all subsequent damage? Are subsequent infections of trypanosomes by the reduviid bugs eliminated? If the answers are affirmative, a vaccine would be a rational approach to this problem. It is not part of my subject, but we should remember that bet- ter housing in this case is an even more rational approach. To go on, which antigens of a parasite induce positive defense mechanisms and which induce suppressor responses? Does treatment of minimal schistosomiasis diminish subsequent immunity? We do not know the answer to this question, but we should, if we are going to treat patients with low grade infections. What disease manifestations result from immune hypersensitivity? We already know that some pathology in schistosomiasis and filariasis is caused by immune reactions, so we 319

should be considering the possibility that immunosuppressive drugs may be useful in treatment of the immunopathology of some parasitic diseases. We need better serodiagnostic tools to use in assessing prevalence of disease and efficacy of drug treatment. We need serodiagnostic tools capable of detecting different strains of parasites and, we hope, some which correlate with parasite burden and can be used to measure intensity of infection. Having implied by these questions that our present store of knowledge about the immunology of parasitic diseases is limited, let me now move on to describe a recent development in immunology that may convince you that it is worth going on. This development, a method of producing monoclonal antibodies using lymphocyte hybridomas, is revolu- tionizing the field and already is spreading to other disciplines.^/ The technology has so many applications to development of serodiagnos- tic tools and vaccines for parasitic diseases that I will take a few minutes to explain it and its potential. It is based on the fact that each antibody-producing cell in the body makes only one kind of anti- body, directed at a single antigenic determinant. If it were possible to isolate a single antibody-producing cell and allow it to grow into a huge colony, that colony would produce monoclonal antibody restricted to one antigenic determinant. This is now being done, and here is how to do it. Two different cells can be fused by a variety of viruses or by polyethylene glycol. Their fusion results in a single cell with two nuclei. Subsequently, the nuclear membrane breaks down, chromosomes divide, and they are randomly distributed into two daughter cells. The daughter cells are hybrids because they contain chromosomes from the two original cells. Kohler and Milstein in England adapted this principle to produce monoclonal antibodies._!()/ For one of the fusing cells , they used a myeloma cell produced experimentally in mice and adapted to culture. It can grow forever. For the second fusing cell, a mouse is immunized with any antigen one wishes (it need not be pure), and the spleen cells from this mouse are fused with the myeloma cell. The cells are then grown in a selective medium. Myeloma cells alone cannot grow in this medium because they lack an enzyme which normal spleen cells possess. Thus, the unfused myeloma cells die, and the unfused spleen cells, not adapted to long-term culture, also die. But the hybrids live on, because the myeloma cell gets the chromosomes necessary to provide the missing enzyme, and, in living on, they im- mortalize the spleen cell's capability to make antibody. The hybrids are grown and then cloned in microtiter plates. The medium from each clone (derived from a single cell) are tested for anti- body to the desired antigen. Positive clones are selected and grown en masse, either in vitro or in vivo. A single mouse carrying these hybrid cells will produce antibody, for example to a lymphocyte marker, 320

amounting to more than all the antibody to this marker made by hundreds of investigators all over the world. Such antibody can have titers in the millions, unheard of before, and it is directed at a single antigen. It can be used to isolate the antigen by affinity chromatography and to develop specific radioimmunoassays. If there are minor differences between different strains of similar parasites, this is clearly the way to detect them. Presently, we are using this system for schistosome antigens and adapting it to detect strain differences in Leishmania. Adaptation of this system to produce monoclonal human antibodies should have amazing therapeutic potential. Before leaving this topic, I would like to point out that this incredibly practical and important discovery stems from basic, non-tar- geted research by scientists in Europe and in the United States who set out to dig deep into the problems of molecular and cell biology and immunology. The long-term goal of any immunologic approach to parasitic- diseases is development of safe and effective vaccines. Recently, there have been important developments in malaria, and vaccines have been produced that successfully immunize monkeys using sporozoites, merozoites, and gametocytes. Furthermore, irradiated larvae have been used successfully to protect cattle against Schistosoma bovis. Dr. Jordan describes these elsewhere in this volume. For development of vaccines, we Tieed to obtain antigens in quantity and to develop adju- vants that can be used in man to enhance the naturally poor effector immune response to parasites. With respect to the need for antigens, there has been a major breakthrough by Dr. Trager at Rockefeller Uni- versity, who has successfully cultivated human malaria merozoites in tissue culture._1l/ Also, Dr. Hiroumi at ILRAD in Kenya is growing African trypanosomes. These are most promising developments. Obtain- ing quantities of antigen from helminths, however, is a greater problem. For example, the schistosomula, the stage of the parasite which makes its odyssey through man, does not divide. We are currently attacking this problem by attempting to define antigen or antigens that elicit an effective immune response by generating monoclonal antibodies. The aim is to determine which of these antibodies to schistosomal antigens can transfer immunity to schistosomula in mice, or mediate eosinophil-depen- dent killing of schistosomula in vitro. These antibodies can then be used to isolate antigens and characterize them. It is not too unreason- able to hope that, if the antigenic determinant is not too large, modern methods of synthesis, or molecular biologic techniques being developed, such as those for production of insulin by bacteria, might be adapted to this problem. As for adjuvants needed for vaccines, considerable work is being carried out to develop adjuvants safe to use in man. Promising pros- pects include muramyl-dipeptide, or MDP, an active ingredient from mycobacteria, and synthetic derivatives of MDP. Other products from mycobacteria and from other organisms such as Nocardia are also being 321

used. Another method for enhancing the immune response is to incorpor- ate antigen and adjuvant into lipid envelopes for liposomes, and this is also being investigated. Hand in hand with this research, it will be necessary to determine which portion of immune responses is enhanced by each adjuvant. The aim here is to be able to enhance a particular cellular or humoral mechanism at will, bypassing suppressor mechanisms. I should like to stress a point that Dr. Krause made yesterday; namely, that the study of parasite immunology will produce, as it already has, information that will be useful in other areas. He gave the eosinophil as an example. In fact, study of parasite immunology has uncovered an unknown and important function of eosinophils; namely, that these cells can kill antibody-coated schistosomula and other larvae. Anthony Butterworth was the first to show that human eosinophils, and other cells, kill antibody-coated schistosomula.jL2,13^/ Under the microscope, one can see the eosinophil, with its granules, plastered up against the larvae. The cells then degranulate, and one can see granules on the surface. In recent studies with Butterworth, and in collaboration with Gerald Gleich, we have shown that damage to larvae is mediated by major basic protein, the main protein found in eosino- philic granules.14/ Until this time, there was no known function for this protein. Probably the most important and exciting area of investigation in immunology today is on the basic control of the immune response. In order to know how to deal with and to alter this complex system, one must know how it works, And this work has produced many surprises. I would like to end with an example of such a surprise. Although this is not parasite immunology, it is directly related, and contains the germ of an important new concept, You may, understandably, be dis- turbed that I should pick a recent finding from the field of bone mar- row transplantation, certainly the ultimate in sophisticated indivi- dual medicine as practiced in superindustrialized countries. But it does bear on our problem, and it illustrates why we must understand control mechanisms. Recently Drs. Reinherz, Rosen, and Schlossman and their colleagues in Boston gave a bone marrow transplant to a woman, with her identical twin sister being the donor. The transplanted cells caused a definite graft versus host reaction in the twin recipient, manifested by skin rash and abnormal liver function. At first, this would seem to be quite impossible, because in identical twins, the transplanted cells should be identical to the patient's own cells, and according to the dogma of histocompatibility, should not react against her own tissues. On analyzing the recipient's blood, it was found that, at the time of the graft versus host reaction, one subset of T lymphocytes was com- pletely missing. This subset had previously been shown to mediate 322

suppression. As the subset of T cells returned, the graft versus host reaction disappeared. Why is this important to us, and what does it mean? It indicates that we normally have autoreactive cells in our systems, and that these are constantly under control by T suppressor cells. If the suppressor cells are absent, the autoreactive cells become manifest. Hence, we must continually have a strong T cell suppressor response to prevent autoimmune disease. This concept has many ramifications. For example, the reason it may be so difficult to mount an attack on cancer cells is that this attack may be suppressed by the same suppressor cells that prevent autoreactions. Furthermore, the parasite, which takes on host antigens, may actually stimulate this suppressor response. This finding has more than theoretical importance, and should certainly be of interest to the pharmaceutical industry. Several anti- lymphocyte sera made commercially for transplantation programs did not work. In fact, some even enhanced graft rejection. As it turns out, some of these sera were directed against the very subset of suppressor T cells we have been discussing. Thus, it is possible that eliminating these cells temporarily with these antisera could greatly enhance immunity against tumors or parasites. As we learn more about the dif- ferent antigens of subsets of lymphocytes, identifying cells with dif- ferent functions, we should be able to favor production of antibodies that can affect certain cell types and thereby modulate the immune response. All of this points out that there may be similarities in the ways by which tumors and parasites evade immune attack. What is more, para- sites appear to be ideal organisms for study and analysis of control mechanisms of the immune response. There is no doubt that much of the information we gain from intensive study of parasite immunology will spin off to directly affect conditions in the industrialized world and in the developing world. Indeed, I would like to leave you with the thought that increases in expenditures by government and the pharmaceu- tical industry on these projects will result in benefits for the health of people here, a substantial fringe benefit indeed for this humani- tarian endeavor. 323

REFERENCES 1. Harrison G: Mosquitoes, Malaria, and Man: A History of the Hostilities Since 1880. New York, EP Dutton, 1978 2. Smithers SR, Terry RJ, Hockley DJ: Host antigens in schistosoraia- sis. Proc Roy Soc Lond B Biol Sci 171:483-494, 1969 3. Sher A, Hall BF, Vadas MA: Acquisition of murine major histocom- patibility complex gene products by schistosomula of Schistosoma mansoni. J Exp Med 148:46-57, 1978 4. Gray AR: Antigenic variation in a strain of T. brucei. J Gen Microbiol 41:195-214, 1965 5. Cross GAM: Antigenic variation in trypanosomes. Am J Trop Med Hyg 26:240-243, 1977 (special supplement) 6. Jayawardena AN, Waksman BH: Suppressor cells in experimental trypanosomiasis. Nature 265:539-54l, 1977 7. Dean DA: Decreased binding of cytotoxic antibody by developing Schistosoma mansoni. Evidence for a surface charge independent of host antigen adsorption and membrane turnover. J Parasitol 63:418-425, 1977 8. Tavares CAP, Scares RC, Coelho PMZ, Gazzinelli, G: Schistosoma mansoni: evidence for a role of serum factors in protecting arti- ficially transformed schistosomula against antibody-mediated kill- ing in vitro. Parasitology 77:225-233, 1978 9. Lymphocyte Hybridomas, Vol 81. Current Topics in Microbiol & Immunol. Edited by F Melchers, M Potter, NL Warner. Springer- Verlag, 1978 10. Kohler G, Milstein C. Continuous cultures of fused cells secret- ing antibody of predefined specificity. Nature 256:495-497, 1975 11. Trager W, Jensen JB: Human malaria parasites in continuous culture. Science 193:673-675, 1976 12. Butterworth AE, Sturrock RF, Houba V, Mahmoud AAF, Sher A, Rees PH: Eosinophils as mediators of antibody-dependent damage to schistosomula. Nature 256:727-729, 1975 324

13. Butterworth AE, David JR, Franks D, Mahmoud AAF, David PH, Sturrock RF. Houba V: Antibody-dependent eosinophil-mediated damage to -'-'-Cr-1abeled schistosomula of Schistosoma mansoni: damage by purified eosinophils. J Exp Med 145:136-150, 1977 14. Butterworth AE, Wassom DL, Gleich GJ, Loegering DA, David JR: Damage to schistosomula of Schistosoma mansoni induced directly by eosinophil major basic protein. J Immunol 122:22l, 1979 15. Reinherz EL, Parkman R, Rappeport J, Rosen FS, Schlossman SF: New Engl J Med 300:106l-1067, 1979 325

MOLECULAR BIOLOGY AND GENETICS AS BASIC APPROACHES TO THE PROBLEMS OF PARASITISM AND SPECIFIC CHEMOTHERAPY Joshua Lederberg I am grateful for this opportunity to offer a few observations on the relationship of fundamental studies in molecular biology and genetics to the world's most urgent health problems. First of all, there are compelling reasons for biologists to put renewed emphasis on parasitic adaptation. There are innumerable ways in which the study of parasitism has illuminated the biology of the host — in a sense we are emulating the evolutionary 'learning' by which the parasites themselves have carved out their ecological niches by taking the most subtle advantages of the habits of their host. Fields such as immunology, hormonal control of development, cell biolo- gy of phagocytosis and endocytosis are rich in such contributions. Studies of the most primitive parasites, bacterial viruses, would alert us to expect that there may be quite specialized adaptations, at the genetic level, between parasite and host — the direct chemical exhibition of DNA sequences, and indirect methods of exposing homolo- gies of parasite and host DNA, will tell us soon enough. These sophis- ticated advances thus enable us to leapfrog fortuitous gambles in the quest for obscure biological relationships. Exploiting similar precedents, we should be looking for phenomena like plasmid-determined, viz. 'infectious' drug resistance, in eukaryo- tic parasites which may pose important complications in chemotherapy of parasitic, as they do in bacterial diseases. To be sure, parasites do not often congregate densely and in mixed company, as bacteria do, and these ecological differences may engender quite different genetic systems. On the other hand, it is hard to believe that parasites that spend a large part of their life cycle within the host cell, in inti- mate contact with its DNA, will have no trace of alien sequences which can have played an important role in accelerating their evolution. Turning now to practical applications, long established methods in microbial breeding should certainly be exploited for production of attenuated live vaccines. Surely, Mycobacterium leprae, though still eluding cultivation in artificial media, can be made to exchange 326

genetic information about its antigens with other mycobacteria more amenable to laboratory manipulation — and the world's armadillos will not breathe easily until we do these studies. Similar approaches have already been applied to enteric pathogens, like V. cholera. New ways of cultivating malarial plasmodia should encourage cognate approaches that could also exploit the host-specificity for virulence of known species. Indeed, these approaches have been technically available for the last 25 years. Somewhat more modern methods of enzymatic analysis of DNA, eventu- ally capped by explicit sequencing of whole (virus-sized) genomes, will be invaluable for working out the taxonomy of difficult groups like leishmania. In fact, the genetic polymorphism of hosts (like people with Hb-A versus Hb-S) is also accessible now to exhibition of DNA differences. These methods having been worked out for visible markers, we can now look forward to comprehensive mapping of the entire human genome. On such a map, it will be easy to locate polymorphisms for genetic susceptibility/resistance to parasitic and infectious diseases, and in turn to the physiological controls that can be exploited for hygienic ends. The most exciting opportunities stem from new technologies of DNA- splicing or genetic engineerig. Enzymatic tools are now available that can cut DNA segments at specified locations, and others that can rejoin and splice segments into viral or plasmid carriers. These in turn can be grown, if necessary in tank lots, in cultures of bacterial or yeast hosts, often with expression of the genetic information that had been spliced in. This new technology is already extensively exploited for the deeper understanding of cellular phenomena. We are just beginning to see efforts at its technical exploitation, which will surely include: • Production of specific parasite antigens, for diagnostic and prophylactic uses. • Manufacture of proteins identical to, or carrying the essential specificities of human immunoglobulins: first for diagnostic and then for therapeutic use (passive immunization and 'sero'- therapy). • Production of specific receptors (like the Duffy factor) that can be expected to interfere with the natural life cycle of a parasite. • Discovery and manufacture of other host factors that interfere with a parasite (in the fashion that high-density lipoprotein lyses trypanosomes.) 327

These extensions have been impeded by the regulatory apparatus which has proliferated in response to hypothetical fears about proli- feration of laboratory DNA hybrids. The unanalyzed concept of patho- genicity is applied in a way that would distinguish between lion and pussycat, as DNA source, in the containment measures imposed on experi- ments transferring such DNA to frogs. Fortunately, the enormous bene- fits of DNA research are being more widely perceived as part of the policy equation; and in many countries we can expect rapid resumption of research of the utmost human utility. At still another level, it must be cautioned that safety of a prospective pharmaceutical can no longer be regarded as a side issue, to be worked out at a late stage in development and validation of a compound. Almost any compound with significant effects on any organism must automatically be suspected of latent human toxicity, and in the present climate must be assumed guilty until proven otherwise. The plan for demonstrating safety of a drug in practical application should be part of the whole R&D design, and in many cases will dominate the overall cost of development. The insights of molecular biology — hav- ing already generated such powerful and valuable tools as the AMES test — may very well also give us more effective policy criteria for pre- dicting effects in human application of suspect agents having irreplace- able therapeutic value. This remark is essentially a call foi estab- lishment of a sorely needed new discipline, a comparative toxicology that can call in all of the tools and wisdom of traditional fields of comparative biology, genetics, and evolution. 328

METHODS OF DRUG ADMINISTRATION John Urquhart This Conference on Pharmaceuticals for Developing Countries addresses means for improving health in areas of the world where economic resources are most limited and health problems are most press- ing. Pharmaceuticals play a key role in both therapy and prevention of disease, and thus constitute key technological resources for improving health. Pharmaceuticals have to compete for limited economic resources both with other means for health improvement and among themselves. Con- sequently, economic and political pressures are strong to minimize the costs of pharmaceuticals. These forces have impact not only on utiliza- tion of today's pharmaceuticals but also on priorities for research and development of tomorrow's pharmaceuticals. One immediate consequence of these economic and political pressures is the move to class pharmaceuticals of recognized value in developing countries as chemical commodities, emphasizing least cost for a total treatment._!/ Naturally, this often means emphasis on least cost per unit quantity, though it is preferable to focus on therapeutic/prophy- lactic outcome and the cost required to achieve that outcome — in those circumstances in which outcome is assessable. One deficiency in viewing pharmaceuticals as chemical commodities stems from the fact that, to be utilized as practical therapy or prophylaxis, pharmaceuti- cals have to possess more than just chemical and pharmacological attri- butes. The additional attributes relate to the methods by which drugs are administered to people. Drug Administration: Basic Considerations From the outset, confused semantics vex discussion of drug administration. Almost everyone, with the exception of the organizers of this Conference, uses the word "drugs" synonymously with pharmaceu- ticals. In fact, a pharmaceutical is one or more pharmacologically active chemical substances compounded with other substances into a form — in pharmaceutical language, a dosage form — that allows the active substance(s) to be administered to a patient. Thus, a single active substance may be prepared in various dosage forms: tablets, 329

capsules, ointments, drops, injectables, and others. The pharmacology of an active substance teaches its therapeutic potential. That potential is translated into practical therapeutic value when the active substance is prepared in a dosage form and the regimen is defined for its use in that form. If the regimen is too com- plex, too productive of unpleasant side effects, or too expensive, the therapeutic potential of the active substance goes unfulfilled because people shun its use. The life-saving hormone, insulin, illustrates the practical value of dosage forms. Everyone recognizes what a problem it is that the only effective dosage form of insulin is the injectable, and how much we could improve both therapy and the quality of life for diabetics if a better dosage form existed for the active substance, insulin. Fur- thermore, many pieces of indirect evidence suggest that the outcome of therapy is degraded because insulin is always administered unphysiologi- cally and often administered unreliably.^/ Man has employed every natural orifice of the body as a portal for drug entry, and sometimes we create orifices temporarily just for that purpose, as for injections of I.V. infusions. My purpose in this paper is to describe opportunities through research for improving therapeu- tics and prophylaxis by improving methods for administering drugs. My thesis is that improvement and simplification go hand in hand, because making something simpler to use makes it more reliable. The reliability of a method of drug administration has both technological and human elements; either or both elements may limit realization of a drug's potential therapeutic value. Use of an unreli- able method to administer a proven drug can be expected to result in an outcome degraded in relation to that attainable when the same drug is reliably administered. A degraded outcome obviously has an adverse effect on any cost/benefit reckoning one might employ in comparing means for improving health. A clear example of how unreliability in drug administration degrades outcome is seen with oral contraceptives: when reliably administered in a closely monitored clinical trial, preg- nancy rates are ca. 0.1 percent per year _3/ but, in actual everyday use, the element of unreliability in unsupervised self-administration degrades this outcome to pregnancy rates of 2 percent 4/ to 6 percent j>/ per year — 20 to 60 fold higher than those pharmacologically attain- able. As Ryder _5/ pointed out in his seminal work in this area, "it ... (is) most important to recognize the extent to which the outcome of use is the joint consequence of method characteristics and user characteristics." To place primary emphasis on creating new methods for administer- ing active substances, and secondary emphasis on creating new agents, reverses an historic gradient in pharmaceutical research. Both acade- mic and industrial research have always focused most intently on the 330

search for new active substances, with methods for their administration taken for granted. Yet, in the past decade, the creation of new methods to administer active substances has attracted growing interest, research, and product development efforts. Analyzing the therapeutic significance of these efforts for developing countries is the focus of this paper. Logistics of Chemotherapy Between the worlds of pharmacology and of therapeutics lies a series of distribution and delivery systems. Their role is to convey an active substance, with its potency preserved, to appropriate drug receptors in patients throughout the world. With suitable packaging, shipping, transfer through customs, local distribution and transport, and with preservation from temperature extremes and excessive moisture, a pharmaceutical can travel, and remain within specified potency, from a plant in Kalamazoo, Michigan, to a dispensary in El Golaa, Tunisia. The last step in the distribu- tion/delivery sequence occurs when the package is opened and a patient contemplates a pharmaceutical whose administration someone has deemed desirable. At that moment the pharmaceutical must pass a final gaunt- let of hope, fear, resignation, indifference, anger, and/or suspicion before it enters the patient. If taken, the dosage form disappears, usually within minutes, as the active substance passes, generally by dissolution or leaching, into soluble form in body fluids. Thence, molecules of active substance proceed by diffusion into the blood, dis- tribution via circulating blood, and diffusion through capillary and perhaps cellular membranes, to reach sites of therapeutic action: the end of a long journey from factory to receptor. Unfortunately, one-time administration of a pharmaceutical rarely suffices, because most active substances are metabolically inactivated and/or excreted within a few hours to a day or two. Thus, the key step of dosage form administration has to be repeated. The practical maxi- mum rate of self-dosing is four times daily. Of course, a highly moti- vated patient or health worker could repeat dosing 10-20 times per hour in the face of great need for continuous presence of agent. There are a few instances of such high frequency dosing, £.£. , in the topical treatment of severe infectious ophthalmias,j>/ but high frequency dosing is exceptional. At the other end of the spectrum, only a few pharmaceu- ticals can provide adequate therapy/prophylaxis on a once- or twice- weekly schedule — for example, very useful prophylactic regimens vs. malaria. Still fewer products can be effective when given every few weeks to months (e^£. , the Depo-Provera R injectable contraceptive and the injectable enanthate of fluphenazine), though one of the limiting features in the long-acting injectables is ambiguity of their dura- tion. Finally, a few immunizations provide adequate protection on a once-a-1ifetime basis. Leafing through any good formulary, however, will demonstrate that the vast majority of useful active substances 331

are in dosage forms that have to be taken l-4 times per day. Repetition of the conceptually simple, widely accepted acts of pill swallowing, eyedrop instillation, ointment application, or injec- tion creates a logistic problem of some magnitude. In Western medicine, the problem is reflected in what is called patient noncompliance. In developing countries, the same problem is a major cause of extensive underutilization of available medicines that is so widely commented upon. Underutilization has both macro— and micro-aspects, and only when both are addressed satisfactorily can the problem be solved. At the macro-1evel, it can stem from such factors as unevenly distributed health care personnel, faulty channels for distributing products, eco- nomically constrained procurement, and so forth. At the micro-1evel, patients overtly accept, but in many instances covertly reject or erratically execute, regimens whose perceived impact on well-being may be substantially at variance with what scientific, medical, or politi- cal authority may teach or wish. My focus here is the micro-1evel prob- lem of noncompliance with scientifically valid regimens for administra- tion of proven drugs. I take it as axiomatic that poor compliance results in degraded outcomes of therapy or prophylaxis; when that is not the case, then good compliers are at a disadvantage of being over- medicated, and the regimen is ipso facto scientifically invalid. An extensive literature, complied and analyzed by Sackett and Haynes,TJ teaches that noncompliance is disappointingly common in Western medicine. It prevails in the face of literacy, technological awareness, health consciousness, and ready access to pharmaceuticals costing no more than consumer products such as cosmetics, cigarettes, and alcoholic beverages. Most studies of Western patients' compliance with regimens of chronic medication show that only about one-half to two-thirds of prescribed amounts are taken. Moreover, it is apparent that timing of self-medication is often erratic, as Kass has shown with his recent work using an ingeniously microcomputerized dispensing moni- tor for eyedrops in treatment of glaucoma.8/ The conclusion is ines- capable that noncompliance is a major worldwide problem with today's pharmaceuticals. It logically follows that improving methods of administering active substances — old or new — to make them far simpler to use could improve therapeutic outcome. The goal is to minimize dependence of therapy upon repetitive acts by patient and/or health workers. More- over, it is essential to accomplish this with no added risk — and pre- ferably with less risk — compared to that incurred using conventional dosage forms. 332

A New Class of Pharmaceuticals A basic strategy in improving on conventional dosage forms has been to develop a wholly new class of pharmaceuticals. These have the characteristic that — instead of rapidly disappearing by dissolution or disintegration — they preserve both form and function, delivering drug at specified rates for an extended and specified duration. Such rate-specified pharmaceuticals are variously called controlled or advanced drug delivery systems, or therapeutic systems.9-13/ The term therapeutic systems will be used here. The physical-chemical princi- ples of membrane-controlled diffusion and osmosis are the basis for many of these developments. A key distinction between a therapeutic system and a conventional pharmaceutical is the manner of specifying strength: for conventional pharmaceuticals, strength is specified only by quantity of contained drug; for therapeutic systems, strength is specified not only by quan- tity but also by the rate at which they deliver their active agent(s) and the duration for which they do so. Specification of strength by rate of drug delivery is noteworthy from a technical point of view, but clearly has to demonstrate its value in practical therapeutic consequences. Therapeutic Rationale of Rate-Specified Pharmaceuticals Two major ideas intersect in assessing the value, in practical therapeutics, of rate-specified drug delivery. The first major idea is that a therapeutic system, precisely bcause it is a rate-specified dos- age form, should maintain uniformity of drug action with little or no side-action throughout the system's functional lifetime. In many ways this point is obvious, yet it has revolutionary implications in pharma- cology, as will be discussed below, and in therapeutics, where reduc- tion of unpleasant side effects minimizes these disincentives to com- pliance. The second major idea is that, by controlling rate, the inter- val between acts of self-medication can be prolonged, in some instances to an unprecedented extent, thus simplifying regimens of chemotherapy. One might note there that the technical capability of specifying deliv- ery rate in vivo, with concomitant reduction in drug level-related side effects, will lift such rate-specified pharmaceuticals above the long and acrimonious debate about bioequivalence inequivalence of various conventional forms of the same active substance.14-16/ Let us examine the principles upon which the second major idea rests. Conventional dosage forms — including sustained-release tab- lets and capsules — deliver their active substance(s) in what Is called first-order fashion: delivery occurs at rates that are highest initially, and then decline steadily thereafter. This pattern approxi- mates that of exponential decay, i.e., first-order kinetics. 333

The result of repetitive dosing with such dosage forms is a saw- tooth pattern of peaks and valleys in the concentrations of agent in blood and tissues. After any particular dose, the concentration of agent at some critical site ultimately becomes too low to be effective; the need then arises to remedicate. The elapsed time from dosing until loss of effect can be prolonged by putting more active substance into the dosage form: the initial rate of release will then be higher, and drug concentration at the critical site will take longer to fall below its minimum effective level; however, the penalty of such prolongation is a more widely fluctuating concentration of drug in blood and tissues, and thus more widely fluctuating effects and side effects. Using first- order dosage forms, one sees the first-order pharmacology of drugs: a time-dependent mixture of side effects and desired effects, with side effects tending to predominate early in the interval following the dose. For some agents, effect and side effect occur at widely different concentrations in blood or tissues, but for many others the difference in these levels is small. With the latter category of active sub- stances, controlled delivery is particularly valuable in separating desired from undesired actions 17/ and, at the same time, extending duration of desired action. Some Principles and Examples of Zero-Order Pharmacology The most astute pharmacologists have now recognized how much conventional pharmacology rests on arbitrary pulse or drench methods of administering active substances. Over the next several years, we can expect to see — under the rubric of steady-state, or zero-order pharma- cology — very substantial rethinking and research on old agents, with discovery of new therapeutic values in some of them. A teaching case is provided by theophylline, largely through work of Weinberger and associates:JL8/ plasma concentrations of this agent can rise a factor of only two above its minimum effective level without causing trouble- some side effects. Controlling delivery of theophylline within this narrow range makes this old agent very much more useful than it has ever been, and competitive with the newest broncho-dilating agents. Because this new value in theophylline is so markedly concentration- dependent, it is delivery rate-dependent. Thus, it is dosage-form dependent, and merely listing theophylline in a formulary will not suf- fice, for in conventional dosage forms this agent is difficult to use, often inefficacious, and occasionally dangerous. Understanding a drug's zero-order pharmacology is to know the rank order of its actions elicited during delivery at a series of constant rates ranging from the lowest rate sufficient to elicit any detectable effect, to the highest rate allowed by common sense and ethics of experimentation. As one progresses stepwise upwards through that range of delivery rates, successively more actions of the drug appear, and sometimes an action in one direction is replaced by its opposite. 334

Table 1 illustrates the rank ordering of actions of three familiar agents: pilocarpine given topically to the eye, and scopolamine and theophylline, both given systemically. As may be inferred from Table 1, zero-order pharmacology teaches a new basis for finding selectivity in drug action, by utilizing zero-order drug delivery to "peel" the top- most action — or top several actions — from the stack of delivery rate—dependent actions, excluding the rest. TABLE 1. THREE EXAMPLES OF ZERO-ORDER PHARMACOLOGY Drug Delivery Rate Pilocarpine a/ Theophylline Scopolamine lowest ocular hypotension bronchodilation miosis myopia ciliary spasm, browache nausea & vomiting tachycardia slight insomnia highest systemic effects, e^jj. , saliva- tion, intesti- nal cramping greater tachy- cardia, cardiac arrhythmias, seizures slight bradycardia inhibition of motion sickness dry mouth inhibition of drug- induced nausea slight tachycardia drowsiness cycloplegia amnesia hallucinations a/Topical, to the eye. This principle is basic in endocrinology: the specificity of a hormone's physiological actions is partly due to chemical structure, and partly due to control of its secretion. This is a physiological lesson of fundamental importance to pharmacology. The demonstrable actions of various hormones include not only their physiological 335

actions but also others that simply never occur unless normal secretory control mechanisms are deranged or the hormone is administered exoge- nously at unphysiological rates or in pulse drench mode (see, for example ref. 19). The lesson should be clear for pharmacology: con- trol complements chemistry in the quest for specificity of drug action. To be sure, there is considerable oversimplification in the sharp distinction I have drawn between first- and zero-order pharmacology. For example, actions of certain agents are known to wax or wane gradual- ly with time, £•£. , tolerance to narcotic analgesics, adrenergically induced "down" regulation of adrenergic receptors, and others. The zero-order rubric cannot accommodate nonstationary actions, but neither can one claim that conventional first-order dosage forms optimally deliver agents with such properties. Instead, agents with nonstation- ary actions merit research specifically to determine what temporal pat- tern of rate-controlled administration is optimal for each agent. Such considerations and exceptions notwithstanding, zero-order delivery can improve therapeutic values of many agents, old and new. Principles of zero-order pharmacology have been reduced to practice in development of an ocular therapeutic system form of the widely-used antiglaucoma agent, pilocarpine.20-23/ This was the first therapeutic system to become a pharmaceutical product. It has a func- tional lifetime of one week, replacing four-times-daily eyedrops; it reduces the total amount of drug required by four-to-eight-fold; it vir- tually eliminated the side-effects of pilocarpine — miosis and induced myopia — that interfere with vision. Conventional, first-order pharma- cology classes pilocarpine as a miotic agent; yet this vision-disturb- ing side-effect has no relation in the steady-state to the drug's desired ocular hypotensive action. A complementary value of automatic, rate-controlled delivery technology is to make it feasible to administer agents that are very rapidly metabolized or excreted. An example is the first utilization of the physiological ovarian hormone, progesterone, in a contraceptive product. 24,25/ Progesterone is so extensively metabolized by liver that it cannot effectively be administered orally save in gram quanti- ties. In contrast to some man-made progestational steroids, such as norethisterone and norgestrel, progesterone has relatively low potency- by-weight and is not a practical agent for parenteral systemic contra- ception. Yet its continuous deployment at controlled rates into the uterine cavity has been achieved for a period of one year, and a recent technological advance has made it possible to extend this duration to three years. Contraception at the 2 percent per year failure rate has thus been achieved with intrauterine delivery of progesterone at a rate of 65 ^ig per day and less, a miniscule increment to the 20-30 mg per day secretion rate of progesterone by the ovarian corpus luteum. In this fashion, the principles of rate-specified delivery have been applied to make practical the pharmaceutical use of a physiological substance that is the gestational hormone of every mammalian species, 336

produced by corpora lutea and, in many species including man, by the placenta as well. In this sense, the toxicologic and teratologic record of the safety of progesterone is the evolutionary history of the class Mammalia. Most endogenous substances that chemically mediate physiological actions are, like progesterone, rapidly metabolized and thus short- lived. Rate-controlled delivery technology opens new opportunities to utilize these uniquely efficacy and safety—verified substances in therapeutics. New Path for Pharmacology The demonstration of multiday, continuous, controlled drug delivery has thus opened up a new path for pharmacology, which is illus- trated by another example from ocular pharmacology in treatment of a parasitic disease. Instead of regarding the eye as the subject of a somewhat arcane specialty, let us regard it as a tissue model for the whole body, having the useful property of allowing direct visual observation. The example is the work of Jones, Anderson, and Fuglsang^26/ which has great potential significance for prevention of blindness due to onchocerciasis (river blindness). They showed that intensity of adverse inflammatory sequelae to the long-recognized microfilaricidal agent, diethylcarbamazine (DEC; HetrazanQy), administered by eyedrop, related to the temporal pattern of drug concentration reaching target tissues. They found distinctly different effects at different rates of DEC administration: at low rates, DEC caused death of microfilariae; at higher rates an undesirable inflammatory response accompanied death of the organisms. Results indicate that zero-order delivery of DEC at the proper rate could lead to gross reduction in microfilarial load without endangering critical tissues by inflammatory responses.16/ Jones and his colleagues further point out that, if ocular manifestations of the disease can be regarded as a model for the disease in other parts of the body, then the ocular reactions they have seen also suggest the value of zero-order systemic delivery of DEC. Human skin appears sufficiently permeable to DEC that zero-order systemic delivery of this agent should be possible via the transdermal route. Langham reported what may be interpreted as a favorable feasibility test of this idea, using a simple DEC lotion.27_/ If, however, it is indeed true that satisfactory therapy requires carefully maintained con- trol, then it would be desirable, as Jones et^ al_. suggest, to develop a specific transdermal therapeutic system for DEC to reduce the systemic load of microfilariae with little or no adverse inflammatory reaction. This work with diethylcarbamazine in onchocerciasis is another case study of how new technology can beget new thinking. In this 337

instance, recognition of the potential application of therapeutic systems technology to controlling DEC administration led Jones and his colleagues to examine, with very simple tools, the zero-order pharmaco- logy of DEC under conditions that simulated continuous, controlled delivery. In fact, their work is the first fruition of the recommenda- tions of a PAHO/WHO meeting on onchocerciasis, held in December 1974,28/ which called for just such studies in recognition of the coming revolu- tion in dosage form technology: "The existing chemotherapeutic agents deserve careful re-evaluation from the standpoint of improving their safety and efficacy by searching for optimal treatment schedules. The schedule dependency of the efficacy of potent chemothera- peutic agents is exemplified in the field of cancer chemo- therapy, in which the careful and painstaking study of dif- ferent combinations of dosage and dosing schedules has made it possible to achieve effective regimens with agents that initially seemed to offer only modest promise. This has been the case, for example, in acute myelogenous leukemia in adults, acute lymphocytic leukemia in children, Burkitt's lymphoma, and Hodgkin's disease. The effort to optimize dosage and dos- ing schedules is quasi-empirical, and its inception need not await the conclusion of extensive studies of mechanism of action nor of animal models of the disease. Instead, the search for optimal schedules can proceed in parallel with supporting studies of drug action in vitro and in animals." "Clinical pharmacologic studies on the schedule dependency of drugs for onchocerciasis should be designed with a view to developing schedules that are possible and practicable in the field. In addition, recent developments, in membrane, polymer, and elastomer technology applied to continuous controlled drug delivery give promise of broad- ening the scope of practical dosage schedules beyond those that have been possible with traditional formulations." Developmental Status of Rate-Specified Dosage Forms Many advances in therapeutic systems technology are still in the developmental phase, but some noteworthy achievements or work in pro- gress include: • One-week-duration ocular therapeutic systems in the form of a thin elliptical film, worn beneath the eyelid for delivery of an antiglaucoma drug;2l-23/ • One- and three-year-duration therapeutic systems, in the form of a T, providing intrauterine delivery of pro- gesterone for contraception with concomitant reduction of 338

menstrual blood loss; 24,25,29,30/ it also appears uniquely well-suited to immediate post-partum insertion,^/ permit- ting institution of an automatic contraceptive regimen at the time of confinement; • Half-year-duration injectable therapeutic system, under development for systemic delivery of a progestational steroid for contraception;32/* • Three-day-duration therapeutic system in the form of an adhesive patch, worn on a few square centimeters of skin surface, for transdermal systemic delivery of a drug for motion- or drug-induced nausea;33-35/ • Multiday-duration therapeutic system of the same adhesive patch, under development for transdermal systemic delivery of an anti-hypertensive drug; • Day-1ong-duration, osmotically actuated tablets, under development for oral administration of many rapidly metabolized/ excreted agents, so that they can be adminis- tered once daily instead of 3-4 or more times daily; 17,36/ • A multiweek-duration ocular therapeutic system under development for antibiotic treatment of the blinding communi- cable ophthalmias and trachoma;37/ • One- and two-week duration miniature osmotic pumps, designed so that they can be loaded with solutions of bio- active agents, and implanted in laboratory animals;^/ these are basic tools for research in zero-order pharmacology in animals;39-43/ • Multiday duration, self-powered, lightweight, and thus portable ,^4_/ or ultralight and thus wearable 45-47/ infusion pumps; these are basic tools for zero-order clini- cal pharmacology and, in tropical medicine, could conceiva- bly find specific field application where none but the parenteral route can do. General technical descriptions of these systems can be found in references 10, 12, and 48, and the entire book by Robinson (ref. 48) is an excellent compendium of diverse efforts now underway in the develop- ment of improved means of administering drugs. Their potential value in therapeutic problems of developing countries is discussed in refer- ences 26, 28, 37, and 49. * This product development effort is cooperatively supported by the Special Programme in Human Reproduction of the World Health Organi- zation and by the National Institute of Child Health and Human Development, U.S. National Institutes of Health. 339

Economic Considerations Unfortunately, the first therapeutic systems products are expensive compared to conventional forms of the same agent. We should hold in abeyance, however, judgment that unfavorable economics foredoom zero-order drug delivery technology to providing products only for an affluent minority. Virtually every major new technology enters the mar- ketplace with product cost at an historic high, for translation of a new technology into its first products involves much breaking of new ground — with costly false starts. At the same time, opportunities for applications of first products are often poorly defined. The electronic pocket calculator illustrates these points. The first scientifically oriented pocket calculator entered the marketplace about seven years ago with a price of $400 — 10-20 times the price of a good slide rule. Today's version of the same product is simpler to use, technologically more advanced, and costs one-eighth as much (Table 2). TABLE 2. COMPARATIVE COSTS OF POCKET CALCULATORS (197l-1978) General Purpose a/ Scientific _b/ 1971 $150 $395 1978 $ 10 S 50 a/Datamath, Texas Instruments. b/HP-35 and its contemporary equivalent, HP-31, Hewlett Packard. The first general-use calculator has undergone a 15-fold price reduc- tion in the same time. There has been great diversification from sin- gle product to a multiplicity, designed to meet diverse needs of users from many callings. Needs of potential users of these products grew and diversified as technology demonstrated what was possible. Slide rules have been driven from the market in this seven-year period. More- over, orders of magnitude more people now use pocket calcuators than ever used slide rules, and there exists not only one, but many markets that even the greatest optimist could not have foreseen in 1971. This seven-year old technological revolution, with its diversifica- tion and cost reduction, did not have to contend with the distraction of government regulation. Such regulation, however well-intended, is anti-innovative. Perhaps one day regulation will be deemed desirable 340

for calculators when someone traces a fallen plane, a fallen bridge, or a miscalculated dose to a malfunctioning pocket calculator, but up to now this industry has advanced its technology with astounding alacrity. Its progress recalls the first decade of antibiotic development, and is humbling to today's pharmaceutical industry, which, in response to safe- ty concerns, has become bureaucracy-ridden both from without and within. Yet as slow and costly as we have allowed ourselves to become in research and development leading to new active substances, it appears possible to achieve significant improvements in dosage form design and methods of drug administration much more rapidly. The economics of large volume production can prevail no less in manufacture of rate- specified pharmaceuticals than they do in other industries generally. Conclusion The therapeutic systems revolution aims at simplifying chemotherapy. To make something simpler to use makes it more reliable. Achieving this simplicity for the user, however, requires technological sophisti- cation in the researcher, the developer, and the producer. It is a complex process to make something simple. Technical developments in controlling administration of active substances have fundamental significance for the science of pharma- cology, both animal and clinical. The prospect of having controlled delivery dosage forms for a wide variety of active substances should increase the priority of research on zero-order pharmacology of many older, basic agents. In the search for new drugs, controlled delivery offers a whole new basis for profiling the actions of compounds in the drug screening process. Concepts of zero-order pharmacology should be introduced into the screening process at the earliest opportunity, for the simple reason that control complements chemistry in the quest for specificity. Until screening is done from the perspective of zero- order pharmacology, therapeutically useful opportunities will be missed. The analogous concept of zero-order toxicology should be examined with great care by academic, industrial, and regulatory people. It may offer a rational path out of the current confusion prevailing in the safety evaluation of drugs given in pulse or drench modes. New methods of drug administration begin to lead one's thinking away from seeing pharmaceuticals merely as chemical commodities. The therapeutic systems revolution brings improved pharmaceuticals that share many attributes of medical devices. These attributes preclude such new pharmaceuticals from being understood and described wholly from the chemical perspective that has dominated the pharmaceutical field. There are, therefore, two strategies for making research yield practical consequences relatively rapidly in the form of improved 341

pharmaceuticals for developing countries. The first is to evaluate zero-order pharmacology of the many agents that have been used and found active, but in one way or another have also proved troublesome. If, as with DEC, theophylline, pilocarpine, scopolamine, and doubtless many other basic agents, controlled delivery can dissociate troublesome activities from useful ones, then an "old" drug is vested with new therapeutic value; if, as with progesterone, controlled delivery can allow practical use of a physiological agent, truly old agents can be vested with new therapeutic value. The second strategy is to determine important parameters of patient compliance in real life, recognizing the wide variety in custom and culture that will influence acceptance and utilization of pharmaceuticals. Such information is no less funda- mental to design of widely beneficial products than are the pharmacolo- gic characteristics of the active substances they convey. It is impossible to over-emphasize the importance of this last point. Yet the problem of noncompliance is so widely ignored that it might, but for its universality, seem a conspiracy of silence among patients, doctors, industry, and government. Major programmatic research support should be provided for study of patient noncompliance — the giant orphan of therapeutics. Up to now, we have led ourselves to believe that the regimen required for effective use of an active substance was inherent in the substance itself. Conventional, first-order pharmacology has perpetu- ated that thought by teaching the distinction between short- and long- acting agents. Yet now we see that the therapeutic systems class of dosage forms has made obsolete that restrictive thinking; the practical ability to conceptualize drug action in terms of zero-order pharmacology provides a new basis for seeking and finding specificity of drug action. The new therapeutic systems class of dosage forms allows unprecedented latitude in tailoring therapeutic regimens to human needs. Improving methods of drug administration, then, is a multifaceted research opportunity for improving "Pharmaceuticals for Developing Countries," for improving quality of drug utilization, and for improv- ing outcomes of their utilization. 342

ACKNOWLEDGEMENTS The author is indebted to many colleagues who have made major contributions, by thought, word, or deed, to the concepts and develop- ments described here: Max Anliker, Ernst Barany, William F. Bayne, Harriet Benson, Pieter Bonsen, Richard G. Buckles, Peter F. Carpenter, S. K. Chandrasekaran, Chandler R. Dawson, Benjamin Eckenhoff, Martin S. Gerstel, William. Griesinger, Takeru Higuchi, Barrie R. Jones, Gerhard Levy, Charles E. Manning, Alan S. Michaels, Howard Palefsky, Bruce B. Pharriss, Virgil A. Place, Robert E. Reiss, Edward E. Schmitt, Jane E. Shaw, John W. Shell, Felix Theeuwes, F. Eugene Yates, Su Il Yum, and last but foremost, Alejandro Zaffaroni, whose vision it was in 1968 to foresee how control could complement chemistry, and created an institu- tion to do something about it. Ms. Constance Mitchell provided invalu- able editorial assistance. 343

REFERENCES 1. Azarnoff DL, et al: The selection of essential drugs. WHO Tech Rept Ser 615. Geneva, World Health Organization, 1977 2. Brownlee M: Normoglycemia as a therapeutic goal in insulin-depen- dent diabetics. Drug Ther (Hosp), July 13-20, 1978 3. Tietze C: Oral and intrauterine contraception: effectiveness and safety. Int J Fertility 13(4):377-384, 1968 4. Ford K: Contraceptive efficacy among married women 15-44 years of age in the United States, 1970-73. Advanced Data, National Center for Health Statistics No 26, April 6, 1978 5. Ryder NB: Contraceptive failure in the United States. Family Planning Persp 5(3):133-142, 1973 6. Lippas J: Continuous irrigation in the treatment of external ocular diseases. Am J Ophthalmol 57:298-305, 1964 7. Sackett DL, Haynes RB: Compliance with Therapeutic Regimens. Baltimore, The Johns Hopkins University Press, 1976 8. Kass MA: Glaucoma report. Noncompliance to ocular therapy. Ann Ophthalmol 10(9):1244-1245 , 1978 9. Zaffaroni A: New approaches to drug administration. Abstract of papers presented at the 31st International Congress of Pharma- ceutical Sciences. Washington, DC, September 7-12, 197l, pp 19-20 10. Yates FE, Benson H, Buckles R, Urquhart J, Zaffaroni A: Engineering development of therapeutic systems: A new class of dosage forms for the controlled delivery of drugs, Advances in Biomedical Engineering. Edited by JHU Brown, JF Dickson. New York, Academic Press, 1975, 5:l-34 11. Zaffaroni A: Therapeutic implications of controlled drug delivery, Proceedings of the Sixth International Congress of Pharmacology. Edited by J Tuomisto, MK Paasonen. Helsinki, Forssan Kirjapaino Oy, 1975, 5:53-61 12. Heilmann K: Therapeutic Systems. Pattern-Specified Drug Delivery: Concept and Development. Stuttgart, Georg Thieme Publishers, 1978 344

13. Zaffaroni A: Industrial Development of Controlled Delivery Drug Systems, International Symposium on Science, Invention, and Social Change. Schenectady, General Electric Co, pp 73-80 14. Barr WH, Adi J: Evaluation of the clinical inequivalency of drug products of varying bioavailability. Ann Clin Res 6 (Suppl):48-56, 1974 15. Corrigan 0I, Timoney RF: Some aspects of the influence of formu- lation on the bioavailability of drugs from solid dosage forms. Ir J Med Sci 143:197-207, 1974 16. Koch-Weser J: Bioavailability of drugs. Parts I and II. N Engl J Med 291:233-237, 503-506, 1974 17. Theeuwes F, Bayne W: Dosage form index: an objective criterion for evaluation of controlled-release drug delivery systems. J Pharm Sci 66:1388-1392, 1977 18. Weinberger M, Hendeles L, Bighley L: Relation of product formula- tion to absorption of oral theophylline. N Engl J Med 299(16):852-857, 1978 19. Urquhart J: Physiological actions of adrenocorticotropic hormone, The Pituitary Gland and Its Neuroendocrine Control, Part 2. Edited by E Knobil, WH Sawyer. Washington, DC, Am Physiol Soc, (Handbook of Physiology, Endocrinology, sect 7, Vol IV), 1974, chapt 27, pp 133-157 20. Shell JHW, Baker RW: Diffusional systems for controlled release of drugs to the eye. Ann Ophthalmol 6:1037-1043, 1974 21. Drance SM, Mitchell DWA, Schulzer M: The duration of action of pilocarpine Ocusert(R}on intracular pressure in man. Can J Ophthalmol 10:450-452, 1975 22. Worthen DM, Zimmerman TJ, Wind CA: An evaluation of the pilocarpine Ocusert(|). Invest Ophthalmol 13:296-299, 1974 23. Urquhart JH: Novel methods of ocular drug delivery, Proceedings of the Sixth International Congress of Pharmacology. Edited by J Tuomisto, MK Paasonen. Helsinki, Forssan Kirjapaino Oy 5:62-72, 1975 345

24. Mishell DR, Martinex-Manautou J: Proceedings of a Symposium on Clinical Experience with the Progesterone Uterine Therapeutic System, Acapulco, October 15-16, 1976. Amsterdam, Excerpta Medica, 1978 25. Pharriss BB: Clinical experience with the intrauterine progesterone contraceptive system. J Reprod Med 20(3):155-165, 1978 26. Jones BR, Anderson J, Fuglsang H: Effects of various concentra- tions of diethylcarbamazine citrate applied as eyedrops in ocular onchocerciasis, and the possibilities of improved therapy from con- tinuous non-pulsed delivery. Br J Ophthalmol 62(7):428-439, 1978 27. Langham ME, Richardson R: The chemotherapy of ocular and systemic onchocerciasis. ARVO Abstracts, Sarasota, April 25-29, 1977. Invest Ophthalmol Vis Sci (suppl) April 1977, p 150 28. Proceedings of the International Symposium on Research and Control of Onchocerciasis in the Western Hemisphere, Washington, November 18-2l, 1974. Pan American Health Organizaton, Pan American Sani- tary Bureau, Scientific Publication No 298 29. Kessler A, Standley CC: Fertility-regulating methods. Recent progress in the WHO program of research in human reproduction. WHO Chronicle 31(5):182-193 , 1977 30. Urquhart J: Prospettive della Contraccezione Basata Sulla Dis- missione Controllata di Ormoni od Altri Farmaci. Atti del 3% Seminario Internazionale il Controllo della Fecondita, Genova, Marzo 3-5, 1977, pp 2l-36 31. Newton J, Harper J, Chan KD: Immediate postplacental insertion of intrauterine contraceptive devices. Lancet ii:272-274, 1977 32. Anonymous: Development of new long-acting systemic agents, WHO Special Program of Research, Development, and Research Training in Human Reproduction, Sixth Annual Report. November 1977, pp 34-36 33. Graybiel A, Knepton J, Shaw J: Prevention of experimental motion sickness by scopolamine absorbed through the skin. Aviat Space Environ Med 47(10):1096-1100, 1976 346

34. Shaw JE, Schmitt LG, McCauley ME, Royal JW: Transdermally administered scopolamine for prevention of motion sickness in a vertical oscillator. Clin Pharmacol Ther 21(l):117, 1977 35. Shaw JE, Chandrasekaran: Controlled topical delivery of drugs for systemic action. Drug Metab Rev 8(2):223-233, 1978 36. Theeuwes F: Elementary osmotic pump. J. Pharm Sci 64:1987-199l, 1975 37. Jones BR: The Bowman Lecture: The prevention of blindness from trachoma. Trans Ophthal Soc UK 95:16-33, 1975 38. Theeuwes F, Yum SI: Principles of the design and operation of generic osmotic pumps for the delivery of semi-solid or liquid drug formulations. Ann Biomed Eng 4:343-353, 1977 39. Wei E, Loh H: Physical dependence on opiate-1ike peptides. Science 193:1262-1263, 1976 40. Bowers CY, Folkers K: Contraception and inhibition of ovulation by minipump infusion of the luteinizing hormone releasing hormone, active analogs and antagbonists. Biochem Biophys Res Comm 72:1003-1007, 1976 41. Pinedo HM, Zaharko DS, Bull J, Chabner BA: The relative contribu- tion of drug concentration and duration of exposure to mouse bone marrow toxicity during continuous methotrexate infusion. Cancer Res 37:445-450, 1977 42. Pettigrew JD, Kasamatsu T: Local perfusion of noradrenaline maintains visual cortical plasticity. Nature 271:76l-763, 1978 43. Siew C, Goldstein DB: Osmotic minipumps for administration of barbital to mice — demonstration of functional tolerance and physical dependence. J Pharm Exp Ther 204(3):54l-546, 1978 44. Robinson LA: Features and benefits of various types of infusion devices. Am J Intraven ther 5(4):30 passim, 1978 45. Wright JE, Buckles RG, Dunn JT, Leeper HM, Yum SI: An elastomeric micrometering system for the controlled administration of drugs. Rubber Chem Technol 50:959-968, 1977 347

46. Leeper HM, Buckles RG, Guittard GV, Lorberbaum MA, Sevilla ER, Yum SI: Role of the elasticity of rubber in the controlled administration of drugs. Rubber Chem Technol 50:969-980, 1977 47. Ho DHW, Brown NS, Benvenuto J, McCredie KB, Buckels D, Freireich EJ: Pharmacologic studies of continuous infusion of arabinosylcytosine by liquid infusion system. Clin Pharmacol Ther 22:37l-374, 1977 48. Chandrasekaran SK, Benson H, Urquhart J: Methods to achieve con- trolled drug delivery: The biomedical engineering approach, Sustained and Controlled Release Drug Delivery Systems. Edited by JR Robinson. New York, Marcel Dekker, 1978, pp 557-593 49. Urquhart J: Controlled drug delivery systems in the treatment of hyperendemic diseases in developing countries, Proceedings of the 1974 International Conference on Systems, Man, and Cybernetics. Sponsored by IEEE Systems, Man, and Cybernetics Society, Dallas, Oct. 2-4, 1974, pp 197-200 348

CLINICAL TRIALS IN DEVELOPING COUNTRIES A. METHODOLOGY OF CLINICAL TRIALS IN DEVELOPING COUNTRIES Alvan R. Feinstein In transplanting modern technology, a developing country must consider its own needs and aspirations, while carefully scrutinizing what the technology has to offer in both advantages and disadvantages. The purpose of the scrutiny is to help planners learn from mistakes of the past, so that advantages can be exploited while disadvantages are avoided. Like all other types of modern technology, controlled clini- cal trials require such a careful evaluation. During the past 30 years, controlled clinical trials have been developed as a powerful method for evaluating agents used therapeuti- cally either to prevent or to remedy disease. The power of the method comes from several major scientific advantages that help reduce or eli- minate the bias with which treatments were formerly compared. The way that these advantages occur can be demonstrated from a review of the basic scheme of a therapeutic comparison, as shown in Figure 1. Treatment A, the active agent under consideration, is con- trasted against Treatment B, which is a "control." It can be an alter- native active agent, a placebo, or no specific treatment. As shown in Figure 2, to compare the effects of the two forms of treatment, we con- trast the outcomes noted in the eligible people, who have been divided into Group A (the people who receive Treatment A) and into Group B (the people who receive Treatment B). In many situations of the past, and sometimes even today, this comparison has been conducted nonconcurrently. When a new treatment has come along, its results have been contrasted against those noted at an earlier time period, using the treatment or treatments available previously. Such nonconcurrent comparisons create the hazard of two potential biases. If Treatment B was used in an earlier era, Group B may have contained people assembled with the old diagnostic standards for eli- gibility, whereas Treatment A may have been tested in a later era, with improved diagnostic methods that allow Group A to include "earlier" or "milder" cases, who would have better outcomes than the Group B patients, even if the two treatments each accomplish nothing. A second source of bias in nonconcurrent comparisons of therapy is that the old 349

FIGURE 1. BASIC STRUCTURE OF A COMPARISON OF THERAPY Treatment A (ictive agent) Treatment B (alternative agent; placebo; no treatment) FIGURE 2. DIVISION AND FOLLOW-UP OF THERAPY X People ^v Group Treatment _. A "A~" Outcome Eligible Group Treatment B B Outcome Treatment B may have been accompanied by ancillary treatment that was either primitive or absent, whereas by the time Treatment A comes along, the ancillary treatment may have been substantially improved. This improvement in ancillary treatment — rather than intrinsic differences between A and B — may then be responsible for differences noted in the results of the two treatments. These two problems can be removed in the modern design of control- led clinical trials, because treatments are compared concurrently with presumably similar diagnostic standards and with similar kinds of ancil- lary therapy. Even with concurrent comparison, however, an important scientific hazard can be produced by the way in which the treatments 350

are assigned. If the therapeutic agents are prescribed according to the ad hoc judgement of a clinician and accepted according to the ad hoc judgement of the recipient, a major bias can arise, as shown in Figure 3. The "good risk" patients may be the predominant members of the group receiving Treatment A, whereas the "poor risk" patients are predominant in Treatment B. With this type of "allocation" bias, Treat- ment A should regularly achieve better results than Treatment B, even if both treatments are ineffectual. FIGURE 3. ALLOCATION BIAS IN SELECTION OF THERAPEUTIC AGENTS. DESPITE CONCURRENT COMPARISON Concurrent Eligible Assignment to People Treatment "Good 'Risks" k"Poor Risks" Treatment ^•Outcome Treatment^. Outcome The reduction or elimination of this problem is a second major advantage of modern clinical trials. They are designed as a planned experiment, as shown in Figure 4, using randomization as a mechanism for allocating treatment in an unbiased way that allows good risks and poor risks to be equally distributed within the limitations of the chance introduced by randomization. FIGURE 4. USE OF RANDOMIZATION FOR "UNBIASED" ALLOCATION OF TREATMENT Eligible People Good Risks & Poor Risks Good Risks & Poor Risks Treatment. Treatment Outcome Outcome 351

Even with randomized assignment of treatment, however, a bias may still arise if observers of the post-therapeutic course, shown in Figure 5, are aware of the treatment received by each patient. Prejudices of the observers may then consciously or subconsciously affect the way out- come is noted and reported. Figure 6, which shows how this problem is avoided — by use of "double blind" observers and/or a "hard endpoint," such as death — is the third major scientific advantage of modern cli- nical trials, in addition to concurrent comparisons and randomized allocation of treatment. FIGURE 5. SOURCE OF POTENTIAL BIAS IN OBSERVATION OF OUTCOME EVENT Eligible People Treatment A Treatment B Outcome Outcome 'Aware' observers FIGURE 6. METHODS OF AVOIDING BIAS IN OBSERVATION OF OUTCOME EVENT Eligible People" Treatment Outcome Treatment B Outcome 'Double BlindI observers and/or 'Hard Endpoint' 352

Yet another scientific effort made in modern clinical trials is to "purify" the groups who enter the trial. As shown in Figure 7, the clinical state of the eligible people is simplified by a series of exclusions that remove many of the complex, heterogeneous features of a human ailment. These exclusions may eliminate people who have particu- larly high or low degrees of clinical severity, or who have the co-mor- bidity of additional diseases beyond the one under study, or who are receiving other medication that may affect the action of the treatments under study, or who give the impression that they may be non-cooperative in complying with instructions for maintaining either the prescribed treatment or the schedule of follow-up visits. The eligible people who are not excluded for any of these reasons then form the group who are admitted to the trial and randomized for therapeutic comparison. FIGURE 7. USE OF EXCLUSIONS TO "PURITY" THE PATIENTS ADMITTED TO THE TRIAL Eligible People Admitted People * k Exclusions due to: Clinical Severity, Co-Morbidity, Co-Medication, or Non-Cooperation Treatment A Treatment B -^•Outcome -^Outcome The last main step, shown in Figure 8, is taken to ensure that results will be statistically impressive. Sample sizes are commonly calculated, either before or during the trial, to provide an ample number of people for admission and study. If the numbers cannot be obtained at a single institution or medical setting, several institu- tions may collaborate to use the same research protocol and to pool their results in a cooperative study. 353

FIGURE 8. DETERMINATION OF STATISTICALLY AMPLE NUMBERS OF PEOPLE Statistically _ Treatment ^ Ample > T- ^ Outcome Eligible Numbers ' A People of >v T i A i • . . i >vTreatment Admitted ^ ^ Outcome People Exclusions t In science, as in politics and economics, there is no such thing as a "free lunch." We should therefore not be surprised to discover that each of these five major advantages is accompanied by some dis- tinctive problems and disadvantages. First, if we rely exclusively on concurrent comparisons of therapy, we reject whatever was learned in past clinical experience. The older data, which often may be quite valuable, are left unused and possibly wasted. Second, if we rely on randomized allocation of thera- py to prevent any baseline imbalances in the treated groups, we leave ourselves susceptible to hazards of the luck of the draw. Working at a P level of .05, we take the chance that one out of every 20 randomi- zations will produce an imbalance of a statistically significant magni- tude. Furthermore, the idea that randomization has equalized distribu- tion of good risks and poor risks may make investigators fail to iden- tify and analyze results separately for good risks and for poor risks. Consequently, we are left in a position analogous to that of having performed clinical trials in groups of mice, dogs, and elephants who are not otherwise identified and who are simply lumped together and called "animals." The person who reads the results might like to know what happened separately in mice, dogs, and elephants — but the investigator reports only what happened in the randomized "animals." The third issue — objectivity in observation — creates several problems. A reliance on hard endpoints as outcome events may lead to an inflation of sample sizes if those hard events do not have a high rate of occurence. Furthermore, if results are expressed mainly in such statements as death rates or remission rates, investigators may neglect the many important clinical phenomena expressed in such soft 354

data as relief of symptoms, ability to function, and other aspects of quality of life. Although double-blind observation is a splendid tech- nique when feasible, it is not always feasible; and furthermore, the double-blind masks are often perforated so that investigators or patients become able to recognize the identity of treatments on the basis of such things as appearance, taste, or certain side effects, such as bradycardia. Besides, belief that the double-blind method will prevent biased observation will often make investigators too complacent to develop suitable methods of "hardening" soft data. As the fourth point in design, "purity" is an excellent attribute for the clinical condition of admitted people, but results found in this "pure" group of patients may not be pertinent for realities of the impure complexity treated in clinical practice. As the fifth main fea- ture of design, statistically significant numbers may be mathematically convincing, but may sometimes lead to an excessive focus on statistical rather than clinical issues in design of the research; and need for large numbers often creates major logistic complexity in acquiring, standardizing, and maintaining the data when multiple institutions must collaborate. Beyond these difficulties, certain additional problems have arisen in the way that controlled clinical trials have been designed and sup- ported by Federal agencies in English-speaking countries in the past two decades. The first problem arises from new methodologic issues involved in what might be called contratrophic therapy. As shown in Table l, the classical trials conducted in the domain of epidemiology have been contrapathic, concerned with primary prevention of disease in healthy people. The classical studies conducted in clinical medicine TABLE 1. DISTINCTIONS OF BASELINE STATES AND OUTCOME EVENTS IN DIFFERENT STUDIES Baseline State Outcome Event Type of Study Healthy Disease Contrapathic (1 Prevention) Clinical Removal of Remedial Manifestation Relief Disease Adverse Contratrophic Progression (2 Prevention) 355

have been remedial, concerned with removal or relief of a clinical manifestation such as pain, dyspnea, or infection. Contratrophic thera- py, however, is concerned with secondary prevention. The aim is to thwart adverse progression of an established disease. Contratrophic goals are illustrated by the use of anticoagulants to prevent thrombo- embolism in patients with myocardial infarction; by the use of lipid- lowering or hypoglycemic agents to prevent vascular complications in people with hyperlipidemia or diabetes mellitus; or by the use of any kind of surgical, radiologic, or chemotherapy to prevent death in patients with cancer. Some of the major therapeutic controversies that rage today have come from difficulties in studying these issues, which make the clinician confront unfamiliar statistical problems in a differ- ent kind of preventive medicine, and which make the epidemiologist con- front unfamiliar complexities in classifying a diseased baseline state. These problems create new challenges in clinical epidemiology — an intellectual domain that is itself still underdeveloped. A second additional problem arises from issues in "potency" of the treatments under investigation. For the sake of the statistical model used in trials, treatments may not be given or evaluated in a clinical- ly suitable manner. For example, inadequate use of double-blind pro- cedures may lead to fixed rather than flexible dosage for a drug that should be given in flexible dosage. In many situations where an inter- mediate target must be regulated — such as level of blood sugar, blood pressure, or serum lipids — results of treatment may not be reported according to success of the regulation. For example, in the celebrated controversy over the UGDP study of diabetes mellitus, one of the points at issue is that treatments were analyzed exclusively according to the therapeutic regimen, but not according to the good, fair, or poor degree with which blood sugar was regulated. Another difficulty has been an inadequate assessment of compliance with oral medication. Thus, when blood pressure has not been lowered by an antihypertensive agent, we cannot be sure whether the agent itself has failed to work or whether the patient has failed to take it. Finally, we have still not developed satisfactory methods for analyzing changes in treatment. If a patient stops taking an assigned treatment after two months, changes to another treatment, and dies four years later, it seems silly to assign his death to the first treatment — and yet this type of analy- sis was used in the UGDP study of diabetes. A third additional problem is deciding the exact role to be served by the "control" treatment. Is it being used to test efficacy, effi- ciency, or effectiveness? For efficacy, we want to know whether the active treatment works. Is it better than nothing? There are two dis- tinctly different types of efficacy to be considered here. For one of them, which is purely pharmaceutical efficacy, the active treatment is compared against a placebo, so that the doctor can contribute his per- sonal iatrotherapy in both active-treated and placebo groups. This type of comparison creates problems in the ethics of randomization and it is also disparate from clinical reality, because doctors do not 356

ordinarily prescribe placebos. The second kind of efficacy, which is purely clinical, would involve a comparison against no treatment. This type of trial is more closely allied to clinical reality, but it cre- ates problems in the ethics of prescribing no treatment as well as in performance of the trial, since patients who are wholly untreated may not return for follow-up examinations. Furthermore, although results are clinically realistic, they may not be readily acceptable in the scientific community. For efficiency, the main question is whether the tested treatment works better than an alternative active agent. If the two treatments are similar in efficacy, so that one is not clearly better than the other, we may need extremely large sample sizes to demonstrate a sub- stantial difference between them or to be reassured that the difference doees not exist. Finally, the question of effectiveness relates to the agent's accomplishment in the community or in clinical practice. Thus, even though a particular anti-hypertensive agent is both efficacious and efficient, it may be ineffective in community usage because people do not continue to take it. The question of effectiveness can seldom be answered with controlled clinical trials, because the trials are usually not large enough or long enough to cover the huge spectrum of people who must be evaluated and followed for protracted periods of time. Consequently, answers to the question of effectiveness will often require not controlled trials, but observational surveys, such as the type of Phase-4 surveillance studies now being developed for new pharmaceutical agents. A fourth additional problem occurs because randomization may be mistakenly regarded as a panacea for all difficulties in evaluating therapy. Despite the many overt and occasionally dramatic advantages of assigning treatment by means of randomization, we cannot rely on it as the major or exclusive method of evaluating all new therapeutic agents that will emerge in an age of proliferating technology. There are too many agents that will need to be appraised; and randomized trials will not be feasible for testing all those agents in all clini- cal situations in which evaluations may be needed. The last additional problem I shall cite arises from the quest for hard data. In a multi-institutional trial, standardization of tests may be sought by sending specimens to a central laboratory — a pro- cedure that adds considerable complexity and expense. Furthermore, information collected in the central laboratory may not always be cogent for the study. What is needed is some intensive planning about what kind of information is necessary, and how it can best be collected. In some instances, an initial effort to standardize facilities or observers of collaborating institutions may be more successful than the apparently simpler, but ultimately less productive, tactic of using central laboratories. The conclusion we can draw from these considerations are expressed 357

in the following eight points. Perhaps the most important point is to be sure that studies are designed to answer practical questions. The presumptive purpose of a therapeutic trial is to evaluate an agent that would be used therapeutically. To evaluate that agent in circumstances that differ from therapeutic usage may have certain scientific and statistical advantages, but clinically results may be meaningless. A second point is to be sure, when the basic protocol is implemented, that implementation of the design does not distort any fundamental goals of the project. For example, although diabetologists wanted to know whether regulation of blood sugar, rather than mere prescription of hypoglycemic therapy,'was a worthwhile activity in patients with diabetes mellitus, the UGDP study was implemented in a way that did not address the basic issue of blood sugar regulation. A third point, which is a corollary of the second, is to choose designs that are suit- able for the major goals of the study, even if those designs are associ- ated with minor scientific imperfections. As several excellent statis- ticians have said, better a somewhat inexact answer to the right ques- tion than an exact answer to the wrong question. Fourth, because of the complexities of multi-institutional collaborations, they should be avoided if single studies can be performed successfully at individual institutions. A fifth point is that remedial endpoints are easier to study than prophylactic ones, particularly if the target events to be prevented have low attack rates, so that very large sample sizes are required for the study. A sixth point is that soft-data endpoints are often more meaningful than hard-data endpoints, particularly if the necessary initial efforts are made to harden soft data. Thus, we might get more clinical value from studying relief of symptoms and restoration of func- tional capacity — if we develop good ways of doing so — than from studying changes in X-rays, changes in blood chemistry, or death. A seventh point is to recognize the potential value of data acquired in ordinary therapeutic situations. This is the information we may have to analyze in circumstances where controlled trials cannot be performed. A worthwhile type of research, therefore, is development of better methods for collecting and using this type of information. Finally, for many important situations, a carefully planned observational survey may be more productive than an experimental trial. As developing countries grasp the opportunity to use the technology of controlled clinical trials, the primary considerations — as in all other modern uses of technology — are to concentrate on choosing worth- while questions and arranging to get worthwhile answers. Technology is an important but a secondary contributor to those primary goals. 358

B. FACILITATION OF CLINICAL TRIALS IN DEVELOPING COUNTRIES Edmund de Maar and J. M. Kofi Ekue The following is a summary of WHO experiences in Chingleput, India and Bamako, Mali in the evaluation of new treatment regimes for leprosy; in N'dola, Zambia, where a WHO clinical investigator and technicians have developed a research infrastructure while investigating anti-schis- tosomal drugs; in Tamale, Ghana, where international staff work in a local hospital testing drugs against onchocerciasis; and in Belem, Brazil where malaria chemotherapy trials are in preparation. From these experiences, broad guidelines are suggested which are not, how- ever, intended as a rigid prescription. Predictions on safety and efficacy under local conditions which derive from trials on affected populations in another region can not be accepted as reliable indicators of future performance. It is highly desirable that the study be carried out where the disease is endemic. A protocol outline stating objectives and procedures will help to determine whether such a study is feasible, taking account of the stage of development of the new agent. Initially, research objectives should be restricted to answering a few simple but meaningful questions. The answers will help in future regional use of the agent for control, and will serve as feedback to the laboratory engaged in continuing research on improved agents. It may almost be mandatory to start with a trial run of the test procedure, using an established drug to detect potential problems concerning avail- ability of scientific and technical expertise, equipment, or reagents. It should be borne in mind that local capability will vary and that in some places it is continuously improving. Problems may exist only in certain neglected areas of clinical research. Reorientation of existing activities towards tropical diseases may be possible. 'Specifics need to be considered for protocol, location, staff, patients, resources, and data. 359

Protocol A protocol should be developed through extensive consultation and coordination, particularly for cooperative clinical trials. It should be drawn up in the region where the disease is endemic, at a workshop attended by a planning group with the broadest possible representation. Its goal is to impart pertinent information, including: the aims of the trial; - a detailed description of activities, especially of such procedures as laboratory tests, biopsies, etc.; report forms that can be made part of the patients' records; - consent forms and warning notices for participating patients, written in the local language; - the treatment allocation procedure and treatment charts, particularly when regimens are reallocated or changed; - the labelling code and emergency code-breaking and referral procedures. An independent measure of patient compliance must be agreed upon. This will avoid branding as ineffectual a compound that the patients have not actually taken, or having the design of the study blamed for results that are difficult to explain. Separate guidelines may be pro- vided for management of adverse reactions, and work manuals for the pharmacist or drug supervisor have proven useful. The investigational drug data brochures should be annexed. A subsequent standardization workshop contributes to the precision of clinical and laboratory results with respect both to the selection of patients and to assessment of response. Byproducts of the workshop are a great sense of fellowship in addition to training and motivation. In cooperative studies, a subsequent meeting of principal investigators to review progress is also highly desirable. Location(s) Although the diseases are widespread, suitable trial sites may be difficult to identify. These should be carefully chosen to avoid con- ditions that might prevent application of suitable scientific standards. It will be necessary to check for the following: Prevalence and incidence of the target disorder and of other endemic diseases. 360

Possibilities of reinfection during the trial or follow-up period. (Can the trial procedures discrimi- nate between suppression with relapse and cure with reinfection?) Local variability in patient tolerance or parasite susceptibility. (Are they likely to affect efficacy and safety?) Capability of the medical infrastructure to support the projected research. Accessibility of reliable, adequate transportation and choice of season for study. (The problem of ready access to study areas, preferably throughout the year is not to be taken lightly. Rural roads described as not likely to be flooded may in practice turn out to be impassable for months because the health worker can- not afford to spend hours stuck in mud.) Attitudes regarding tendency to please, security (be wary of drug leakage for alternative uses of the drug), attitudes toward trials of new agents, investigational invasion of privacy, (£•£., blood tests, stool collec- tion, and post mortems). Check absences for such occa- sions as religious holidays, harvests, sickness in the family, funerals. When an appropriate site has been identified, it is appropriate to: - Explore the will to investigate a disease problem of local importance and global consequences. - Make a contract with national and local authorities, including national research councils, so that any agree- ment will extend throughout the administration from the Ministry of Health to the local authorities, and desig- nate the project as national or local with no emphasis on internationality. - Check the necessary clearances for investigating a new agent; both international and local permits may be needed in the country of origin of the drug or vaccine, from a review body, a manufacturer, or from an authority. 361

- Determine whether cooperative arrangements can be made; make full use of all locally available experiences in relevant medical research; and look for a twinning arrangement with a group with proven experience in clinical trials, particularly for clinical pharmacology, pharmacokinetics and training. - - Obtain all possible cooperation, not only from partici- pants in a trial, but also from all inhabitants of the area under study. When the community understands what is going on, work is likely to be facilitated. Neigh- bors may even see to it that patients turn up on days of study. - Consult such local health dignitaries as the traditional medicine man and the dispenser or pharmacist; their cooperation and understanding of the clinical trial should be sought. Their displeasure may jeopardize the whole study. Staff It is essential to find a competent, mature and enthusiastic principal investigator and to ascertain his (her) availability for the duration of the study. Recommended by senior health, hospital or facul- ty authorities, the principal investigator should ideally be a member of the local medical profession and, as such, acceptable to government health agencies. Dependable and, where possible, fully paid by the project, he (she) should be knowledgeable about clinical trials or will- ing to learn, and should have been a member of the protocol working group. The principal investigator(s) must be prepared to do most of the work, the volume of which should be planned with that condition in mind. One should be wary of too eager a delegation of supervision. The trial group may include other medical and supporting personnel, such as physicians, nurses, secretaries, a pharmacist or drug super- visor, a financial or administrative officer, a statistician, and a clinical pathologist who are organized and prepared to substitute for each other in case of sudden unforeseen absences. Even the best moti- vated groups may experience the full range of causes of absences from accidents to unexpected departures for sabbatical leave. Clinical trials offer to young physicians training in planning and executing a time-1imited project, and the opportunity to collaborate with senior scientists. Such an experience may well be the first step on the lad- der to a research career. A coordinator, familiar to the group and not alien to their cul- ture, who adheres to high scientific standards and can appreciate local 362

problems is very useful for the necessary liaison with local groups in other locations, and with the sponsor of the trial. To promote self- reliance, national counterparts should be appointed for any non-1ocal, temporary staff. Patients There should be agreement on the population from which the sample will be drawn, and on the method of sampling of research subjects as regards the intensity of the disease, concurrent diseases, sex, age and ethnic origin. In addition to careful records of name, the household, tribe, domicile, and occupation, photographs may be useful to confirm identification when appropriate. A calendar for clinic visits should be agreed upon, and special clinic hours arranged for research patients. By reducing transportation time, satellite centers can promote recruit- ment of patients and encourage participation by all interested persons in the trial area. These centers should also offer medical care for conditions other than the disease of interest to the study. Care should be taken during selection of a study population to avoid groups who have easy access to conventional medical treatment, for they might obtain treatment which could interfere with the clinical trial. In selection of school children for a clinical trial of a schis- tosomicide, for example, urban schools or rural schools located near a clinic are to be avoided, if possible. The drop-out rate is usually minimal from a study which is well conceived, planned, and explained to the subject population. If a fol- low-up is needed of people in scattered villages one should try to assess mobility as nearly as possible before work is started. Mobility is usually underestimated, and many participants may therefore not be traceable. Ethical Clearance Verbal or written informed and meaningful consent must be obtained from each patient. The methods for explaining safety issues must be predetermined. Handing out consent forms to people who may at best be semi-1iterate would not suffice. One may have to seek additional con- sent from the traditional "head" of the unit as well as from parents and from the head teacher if school children are to give specimens of blood or urine. The consent of a husband for participation of his wife should also be obtained in some areas. Withdrawal of subjects several weeks or months after a trial has been in progress is the penalty for neglect of these precautions. Projects should be cleared by a locally approved, independent body practicing the highest standards of protection of the rights of the 363

individual. Such a body should be created if it does not exist. When presenting the study to that body, the information to be conveyed to potential subjects should be described in detail and should include: aims of the research project; - expected duration of the study; alternative methods of treatment of the target disease; - study procedures which are experimental; - any known immediate or long-term risk or possible dis- comfort to which subjects are liable; benefits anticipated for subjects or others; the degree of freedom of each subject to withdraw from the study at any time; confidentiality of any information relating to patients should obtained during the course of the study. Explanations should be included as to the manner (oral or written) in which this information will be conveyed, and names and status of staff members who transmit this information to potential subjects and who ascertain that it has been understood and that consent is freely given. The written consent form should be appended, if applicable. Any special incentives or treatment services to be given to sub- jects in exchange for their participation should be indicated. Resources 1. Drugs Test medications should never be smuggled into any country; a worthy scientific purpose is no excuse. Research drugs are new chemicals which cannot be classified by customs. A full formula should be placed on the tear-off label and the problem of unrealistic value assessments should be discussed with the manufacturer. The method for obtaining post-trial follow-up treatment supplies should be considered in advance. 2. Samples Responsibility must be assigned for carrying out each particular test of the laboratory work. If assays are to be run elsewhere, and methods of preservation are available, samples should be shipped on a regular schedule. A specimen identification form can be helpful. A laboratory report form links patients and laboratory data. The local laws on export of biological materials should be care- fully checked. It is "medical colonialism" to take samples out of a 364

country, leaving no further trace of that export except a publication many years later from a laboratory in another part of the world. 3. Maintenance Availability of services for all laboratory equipment should be checked. 4. Time A realistic estimate of the duration of the study is needed with starting and completion dates. A flow diagram is useful. The rate of expected intake can be predicted by continuous (weekly/ monthly) plotting of actual intake. Too rapid early intake, or too slow or unsteady intake can lead to incomplete data collection, missed visits and general disappointment of staff. Adding more field stations could increase output without disrupting a study. Consider the conse- quences of delays. 5. Funds The financial officer must determine the totals required and obtain the funds. Data Multi-center, short-term trials rapidly produce large amounts of data suitable for tabulation. For studies of longer duration, or for those involving many patients, it is useful to fix definite times for interim analysis, _e.£., when 25 percent of the sample population has completed treatment. Continuous tabulation may be useful to detect trends signifying toxicity. Whether these results should be supplied to all investigators (at the risk of engendering biases on their part) should be carefully considered. Mechanized methods for data collection, storage, retrieval, and analysis are desirable to: - - promote accuracy and thus ensure acceptability of the information to others; promote close surveillance of the project with quality checks as data are generated; keep up with the rate of data collection and thus ensure that the sponsor receives all data as they become available; obtain significant savings in time and effort. Before beginning, there should be agreement on ownership and con- fidentiality of data, on who is to analyze interim and final results and where, and on publication policy (e^£. acknowledgements), including public announcements. 365

Conclusion If these specific guidelines are observed, clinical studies carried out in tropical countries on the diseases endemic in those countries can proceed as smoothly as trials carried out in more developed, temper- ate countries; the results obtained will be just as reliable. Acknowledgements The authors wish to acknowledge the review of the text and other valuable contributions to this presentation by Dr. C. J. Canfield, Walter Reed Army Institute of Research, Washington, D.C., U.S.A., Dr. B. 0. L. Duke, WHO Headquarters, Dr. V. Eyakuze, WHO Brazzaville, Dr. C. 0. S. Iyer, Chingleput, India, Dr. J. E. Jefferies, Pfizer, Inc., New York, U.S.A., Dr. 0. 0. Kale, University College Hospital, Ibadan, Nigeria, Dr. A. 0. Lucas, WHO Headquarters, Professor P. D. Marsden, Brasilia, Dr. A. Walker, CIBA-GEIBY, Basle, Dr. Waters, Kuala Lumpur, Malaysia, and Dr. D. H. G. Wegner, Bayer A.G., Wuppertal, Federal Republic of Germany. They also would like to acknowledge with thanks the contribution and help of the scientific editor, Dr. V. Weinstein de Jaque. 366

OPPORTUNITIES FOR RESEARCH - AMERICAN INDUSTRY Pedro Cuatrecasas If industrialized nations are to contribute to the solution of the health plights of developing nations, they must apply the strengths and resources which they have most uniquely developed. Because for the phar- maceutical industry this resource is clearly that of drug discovery and innovation, the process by which this occurs must be clearly understood if one is to integrate new or additional challenges in a meaningful, productive way. As this essay will try to demonstrate, applications to specialized diseases must be an integral part of the discovery process in general if we are in fact to use industry's existing and proved assets and capabilities rather than to artificially create well-meaning programs that may turn out to be wasteful and unrewarding. Insofar as opportunities for research in American pharmaceutical laboratories can- not easily be segregated along therapeutic lines or disease states, opportunities and problems related to diseases of developing nations must be viewed and examined in the context of the entire research base and structure of this industry.jY This essay will therefore address primarily the general features of research and development (R&D) which affect the industry's overall research opportunities. In the past forty years, the ethical pharmaceutical industry has been extraordinarily successful in discovering new drugs through its primary emphasis on innovative research. The value of innovative research has been the basic, unique characteristic of the industry which is responsible for its existence and without which it cannot be expected to survive in the future. The contributions of new therapies to the control of diseases, and the impact on the health care systems of our societies, have been profound. Thus, the discovery of drugs and their expeditious evaluation is industry's most important resource. It is clear that there is still great need in all countries for new remedies for treatment and control of serious health maladies. Drugs and vaccines are the most cost-effective forms of medical therapy, and they constitute our major technology for further reduction of costs, as well as the further curtailment of disease. Public policy related to health care must therefore be formulated in the context of the poten- tial contributions which can be derived from new drugs, and adequate 367

thought must be given to ways in which drug discovery can be encouraged, nurtured, and promoted. If it is accepted as public policy that new drugs are desirable and needed, serious attention must be given to the nature of the drug discovery process. To deal with this issue properly, it is necessary to understand the extreme complexities, difficulties, and uncertainties involved in developing new drugs in the present era. The resources and requirements for discovering drugs for diseases of developing countries are, in my opinion, inseparable from those of drug discovery in general. It is naive, if not counterproductive, to believe that it is possible to establish special drug development programs successfully or to direct policies to particular disease states 2_f out of context or with- out addressing some fundamental problems of drug innovation in today's world. If American industry is to contribute, it must do so within the same framework of technical, regulatory, and economic constraints that restrict all its other, current innovative pursuits. I firmly believe that, scientifically and technically, opportuni- ties for discovering and developing new therapies for diseases in all countries are greater today than ever before. For the scientist, very exciting opportunities exist on the basis of the substantial progress of the past 20 years in fundamental biological and medical sciences._3/ The obvious deficiencies and needs of medical practice also provide strong challenges and incentives for delivery of health care. Despite this, current economic, political, and regulatory climates, and their impact on public opinion, are in my view seriously blunting drug inno- vation and threatening to deprive society of the benefit of potential new drugs. In my view, this is the extreme paradox of today: at a time when the potentials of science and technology are greatest, so also are the man-made restrictions, disincentives, frustrations, and discouragements. Because of the overwhelming significance of con- straints that handicap all drug R&D, and because of my belief that much of this constraint results from a lack of understanding of the factors involved in drug research, I will address here some of these general principles rather than directing my comments to specific programs where there is research opportunity in a defined, limited class of disease states. I will try to develop the theme that novel drug therapies will almost certainly arise primarily out of basic R&D programs of the phar- maceutical industry. Success can best be achieved through constructive, collaborative relationships with government and academic institutions. A mutual spirit of understanding and cooperation will be required in order to encourage drug discovery. Comparative Sources of New Drugs In trying to anticipate the possible future sources of new drug 368

therapies, it is instructive to examine historical perspectives and trends, since this is the only available data on which objective, pre- dictive judgments might be made with any degree of accuracy. Recent studies indicate that during the past 35 years the over- whelming majority of significant new drug introductions, defined as New Chemical Entities (NCEs), were discovered and introduced by the pharma- ceutical industry. For example, Schwartzman kj estimates that industry accounted for 91 percent of all such discoveries and introductions in the United States in the period 1960-1969 (Table l). Furthermore, the trends since 1935 have been steadily shifting more and more to the source within industry ^4 ,_5/ as demonstrated by separate studies by Schnee, Seife, and Schwartzman (summarized in ref. 4). TABLE 1. PERCENTAGE OF NEW CHEMICAL ENTITIES (NCE) DISCOVERED AND INTRODUCED IN THE U.S.A. BY THE PHARMACEUTICAL INDUSTRY l/ Period in Which Drugs were Introduced Source 1950-1959 1960-1969 Total 100% 100% Industry 86% 91Z Other 14% 9% It is important to note that these data refer to discovery as well as the development of new drugs ,4^_5,^/ since there are prevalent common public misconceptions that see industry R&D as merely seizing upon or developing (in a routine way) basic discoveries made in academic centers. The fact that in modern times industry has been society's main instru- ment of drug discovery must be recognized, not to exalt industry, but to understand better the forces and factors with which we must work if we are to maximize future discovery._7/ As will be discussed later, drug R&D has become increasingly complex, and it has unique features toward which industry has directed its attention. Academic and government laboratories, on the other hand, 369

have been relatively unproductive in this area (with the exception of vaccines) despite the increased allocations of resources for medical research that now far outweigh those of industry (Table 2). As govern- ment-sponsored research, through its research grant and contract system, TABLE 2. NATIONAL SUPPORT FOR PERFORMANCE OF HEALTH-RELATED RESEARCH BY SOURCE OF FUNDS ($ Millions) l/ Institution 1960 1965 1970 1972 Total 798 l,715 2,499 3,102 Government 471 l,229 l,740 2,223 Private, academic 121 158 193 211 Industry 206 328 566 668 Industry % of total (26%) (19%) (23%) (22%) has proliferated and matured over the past forty years, the proportion of new drugs contributed by this sector has declined steadily. This is not meant to belittle or downgrade the enormous contributions of those institutions to advancement of fundamental knowledge, but to indicate that their fundamental orientation is such that it is unlikely that they will be major sources of applied research and of new drugs for the future. With some exceptions, academic research does not have the orientation, temperament or organization required to mount successful programs for drug discovery and development. In addition, non-industri- al research is not equipped to deal with the complex regulatory pro- cesses which are in today's world intrinsic to and inseparable from the process of drug discovery itself. However, as will be described later, these institutions can contribute along with industry in important ways to specific programs of drug development. But, in my opinion, it is the American pharmaceutical industry which, because of its special interdisciplinary scientific and technical skills, organizational strengths and economies of size is best suited to carry the principal initiative in this area of innovation and development. 370

Threats to the Drug Discovery Process Several factors encourage the pharmaceutical industry to decrease its commitments to innovative research. Spiraling costs of R&D (esti- mated to amount to an overall average of $25 to $40 million per NCE), 10-12 years required to bring a new entity into the market, risks and uncertainties of the discovery process itself, proliferating and restraining regulatory machinery and governmental economic policies which do not favor profits by the innovator have all worked to seri- ously reduce the appeal of drug discovery and return on investment of research. It has been estimated ^/ that the return is now 3 percent on an industry average, which is not a sufficiently high level for prudent investments relative to other alternatives (£•£. , about 10 to 12 per- cent for other industries). There is little question that drug com- panies are now being forced to turn away from competition in innovation ^i • e^., new drug discovery) to competition in pricing, alternate formu- lations and presentations, defensive R&D, etc. These reactions are being largely fostered by public policy, which has been adversely shap- ing public opinion, perhaps not deliberately but nevertheless with negative impact. There is little doubt that spiraling costs of industrial drug R&D, bolstered by staggering costs of compliance with proliferating regula- tory requirements, have markedly shifted distribution of resources away from substantive, creative activities toward less productive develop- mental efforts. Even with increasing R&D budgets, there has slowly emerged an important shift of attention toward relatively nonproductive, bureaucratic tasks which are required for survival. Costs of adapting to increasing regulatory requirements and standards, such as Good Manu- facturing Practices (GMPs), Good Laboratory Practices (GLPs), and Cli- nical Practices, among others, have been immense. The net result has clearly been a substantial decrease in efforts devoted to basic research programs which on the surface may not be too visible because it repre- sents a re-allocation of funds within large total R&D budgets.ji/ In this context, laboratories have been forced to restrict themselves to more limited fields of research, and to abandon or avoid programs that from the outset do not offer attractive economic rewards, regardless of their scientific merit or appeal. This is one of the reasons why many laboratories have curtailed research in tropical diseases. It is notable that despite this, several research-based pharmaceutical com- panies have maintained large programs directed to diseases of the Third World, a situation that may be judged by some dispassionate observers as being economically foolhardy. The Wellcome Foundation, for example, is spending at least one million dollars a year in such research.^/ It is important to note that even if drug research by non-indus- trial laboratories need not be concerned with economic rewards or oppor- tunities, it must nevertheless still conform to regulatory requirements if it is to seriously participate in drug discovery and development. This will involve enormous problems since in addition to bureaucratic 371

restrictions and standardization requirements, for which academia has a strong distaste, it will be necessary to divert funds away from innovative research. A large proportion of the requirements, even if based on a rational principle, stifle creativity. It is unlikely that academia can tolerate standardizations which in many areas are incom- patible with originality.10/ The decreased allocation of efforts and resources to substantive research by industry, although understandable, is alarming. New drugs being introduced today are not today's drugs or discoveries, but those of ten years ago. Today's decreased efforts and discoveries will only be evident ten years from now, when it could be too late to do much about it. We see now a reluctance to tackle important but difficult research programs that have of necessity long-range objectives. Such programs are too large, expensive, and risky. The pressures are for shorter term, more certain but less revolutionary contributions. There are pressures that discourage bold programs and aggressive pursuit of tough problems, even though the fundamental scientific knowledge is available to point to a sound rationale and potential feasibility. The tendency is to be super-cautious, to avoid possible research failures or anticipated demands or denials by regulatory agencies. Potentially useful compounds, for example, are being dropped from further develop- ment because an in vitro test shows potential mutagenicity, or because of a suggestion of addictive potential. To me, this is the drug lag of substance. It is an absolute rather than a relative lag. Many potentially useful drugs are not being examined (or discovered) at all, and society may ultimately be denied discoveries within our technological capacity to achieve. If this is the case even for important breakthrough drugs in our industri- alized nation, where good potential for economic rewards exists, is it any wonder that few deliberate programs exist to pursue possibly non- profitable "orphan" drugs or those for disease of other nations where marketing factors are unpredictable? Need for a Proper Climate for Industrial Drug Innovation It is widely recognized that over the past ten years we have seen a dwindling rate of significant new drug introductions,Ll/ which is especially severe if it is judged relative to the enormously increased research expenditures (by all sources) and basic scientific progress which have occurred during this same period of time. It must thus be recognized that it is not just tropical diseases or rare disorders that have apparently been "neglected" in terms of the supply of new, improved drugs. The reduced productivity is, instead, a reflection of an affliction which has affected drug discovery in general. The challenge now in new drug discovery is how to overcome the obvious constraints and substantially increase efforts and resources 372

allocated to basic research programs in this industry. Industry must be encouraged to exploit in its research the vast wealth of fundamental scientific information which has been accumulating, and to better inte- grate its strengths with complementary talents in the academic communi- ty. To do this, the potential economic rewards for success must be assured; incentives and encouragement must be such that competition in drug discovery becomes the dominant force. A healthy, vigorous, and economically rewarding research community within the industry will result, I am sure, in substantive contributions in many therapeutic fields, including those of the developing nations. Pharmaceutical research is highly unique, unpredictable in the direction of its results and characterized by serendipity. As a matter of principle, the best way to assure discoveries in some areas is to encourage excellent research in all areas. New therapies for parasitic diseases, for example, are as likely to result from a program initially directed to the discovery of an antimetabolite, an anti-bacterial, or even an anti-hypertensive as from a deliberate, rational program for a specific parasitic disease. History is replete with such examples, which despite their appearance of being totally accidental, in fact require scientists with intuition and insights based on experience, and broadly based and well-organized programs. In today's scientific arena this is even more likely to be the case than ever before because of the increasing interdependencies and overlapping nature of scientific dis- ciplines. Certainly, specific programs in therapeutic classes must exist, but they must all be closely integrated. Regulatory agencies, whose task it is to protect against hazards and the regulation of new products, should somehow be balanced in their judgment by public policy mandates which acknowledge the cost-effective- ness of this type of health care technology and the need and value of new therapies. This will encourage innovation. Drug development should proceed in a regulatory environment dominated by encouragement and optimism, not suspicion, adversarial postures or hostility. The fundamental role of the pharmaceutical industry in drug discovery needs to be recognized. The need to recover enormous invest- ments in R&D should be acknowledged and fostered. The nature of its research technology, which differs in important ways from that of other -industries, must be better understood. The rate and nature of success must be examined for the industry as a whole, and not taken out of context to assess its value in selected areas. For example, most new drugs will have poor sales and low return,_10/ which is tolerable if those commercially very successful drugs are accepted and allowed. It must be appreciated that it is those few, highly successful drugs that must recover the costs of R&D of all the failures (which never make it to the market) as well as those of low profit or for relatively rare diseases. If these few major successes are unfairly singled out for attack on the basis of cost (or pricing) or possible monopolistic prac- tices without regard to the totality of R&D requirements and rate of 373

returns on investments, funds for investment in future risky ventures can be seriously compromised. These few highly successful and profit- able drug products may make possible subsequent discovery of new drugs, some of which would undoubtedly be for rare but important diseases. This may in effect be the means by which the consumer can pay for sup- porting future drug discoveries without deliberately re-distributing resources through the government. Programs such as MAC and the policy of generic substitution, product liability and proliferating regulatory requirements further complicate the tenuous profitability structure for R&D investment. The focus of federal attention should shift away from excessive and narrow considerations of pricing (which, to possibly save consumers a few dollars today, may result in loss of new drugs which could result in truly substantial economic and health savings tomorrow) and of misguided interpretations of monopoly ascribed to a few novel, successful agents.4^_5/ Instead, efforts should be re-directed to ser- ious threats to future technical innovations. Influence should also be brought to bear on other countries, since many of these, even some high- ly industrialized countries without significant input in innovational R&D, apply unrealistic controls and disregard patent conventions. The American consumer may be unfairly asked to carry the burden of dispro- portionately supporting new technologies in the area of new drugs if widespread duplications and imitations are allowed. Characteristics of Industrial Drug R&D In an era of enormous potential and great need for discovery, but also of abundant disincentives, it is particularly important that aca- demia, government, and even industry be well-informed about the special skills and technical resources required for successful, innovative drug development. It is sometimes said that basic research and discoveries should be left to academic and government research institutions, and that indus- try should concentrate on "applying" or "developing" those "discoveries" in the form of new drugs. Such views totally ignore the modern history and nature of drug discovery. Although it may be that basic discoveries and new principles in physics, for example, can be readily translated into applied processes or products, it simply does not work out that way in the biological and pharmaceutical research field. As noted earlier - (Table l), in recent times the vast majority of new drugs have been both discovered and developed in private laboratories. It is very likely that this will continue to be the case, perhaps even more so in the future, as discovery and development become more regulated, sophis- ticated, and complex. The fact is that drug development today is an extremely complex and convoluted process that requires many years of persistent collaboration between chemists, biologists, physicians, and other experts in a highly empiric fashion. It involves continual accumulation of ideas and information in small incremental steps by teamwork which bridges many different fields. A new concept in a 374

biological process, a new insight into the molecular events of a pathologic state, or even complete knowledge of the specific genetic basis of a disease, despite their scientific importance are not enough to "develop" a drug. These may provide the stimuli or theoretical basis for specific, concerted efforts to discover a particular type of drug, or they may allow establishment of improved biological models or testing systems, but there are such enormous gaps in our knowledge of theoretical biology and chemistry that the subsequent process of obtain- ing a useful drug is highly empiric, risky, and unpredictable. We have known, for example, the molecular basis of many genetic diseases and abnormalities of the hemoglobin structure (£•£., in sickle cell anemia) for some years, but these basic discoveries have not yet led to new drugs for the correction or treatment of many of these disease processes. Even when a specific chemical lead is available, innumerable things remain to be done (e^j£« , synthesis of all reasonable analogs for comparative structure-activity studies in as many tests as possible) and many obstacles (e_.£. , numerous toxicology studies, metabolism, absorption, disposition, formulation, ease of synthesis, etc.) to be met even before testing for efficacy in man. The first introduction into humans may only serve to gain information which the clinician will convey back to the chemist and pharmacologist in order to improve the parameters of the drug under study. This discovery and development process must be continuous, intensive, and must be sustained over many years before the drug is "discovered." At many stages it must abide by complex regulatory requirements and obligations, even in the early dis- covery phases. Identification of candidate drugs and safety and effi- cacy evaluation involve an interplay which makes it difficult to iden- tify a particular moment of discovery. It is very unlikely that in today's world such a drug would be "discovered" independently in an academic setting because the organization to do this does not exist there. In this setting it is exceedingly difficult and most often unproductive for an industrial laboratory to "take over" or assume a specific lead or discovery that has arisen elsewhere (unless it is far advanced in efficacy testing), regardless of whether this comes from an academic or another industrial laboratory. Industrial laboratories involved seriously in drug discovery must have large groups of specialists in many different disciplines, and they must work together collectively and efficiently over protracted time periods. Even within disciplines a broad spectrum of specialists must be present to obtain optimal scientific cross-fertilization and to exploit various leads and ideas simultaneously in many systems. Close communication, scientific curiosity, motivation for achievement and excellence, astuteness of observation and cognitive and technical skills must all be present to seize on the unexpected. Often work in one area will lead to a "spin off" in another area. In our company, the extraordinary discoveries by Dr. George H. Hitchings in one parti- cular basic scientific field (_i._e. , purine and pyrimidine biochemistry) 375

led to the discovery of drugs important in treatment of such diverse diseases as malaria, gout, bacterial infections and cancer, and for kid- ney transplantation, yet these were not diseases which were originally "targeted" for attack by the basic scientific work which was very broadly based in biochemistry. The principal "targeting" cones from being consumed in the drive to have one's basic work lead to the dis- covery of a new drug, in any field, and from having the organization, resources, and motivation to assure the realization of a final product. Industrial laboratories of research-based companies must undertake substantive basic or exploratory research if they are to have things to "develop" in the future. The distinctions betwen basic and applied research in drug discovery are meaningless, for these merge ever so closely, and they go hand-in-hand. Some of the most important "drug discoveries," such as the contributions of Pasteur, Ehrlich, Fleming, and Domagk would today be classified as "applied" research. The contem- porary "applied" research of industry is equally challenging and meri- torious, and in addition it continues to contribute important fundamen- tal knowledge to the general scientific base. Science in industry is not bureaucratic or rigid, as some might believe, but science there is the same as elsewhere except that collective, multidisciplinary efforts are the rule, and there is a compulsive drive for discovering new thera- pies. Academic labs, in contrast, are highly discipline-oriented, and the obsession with individual discovery, achievement, public recogni- tion, and "advancement" further restrict interdisciplinary approaches, especially over long periods. The primary objective and underlying motivational forces are excellence in basic discovery, for its own sake, publication, and generation of research funds to keep the cycle going. In addition, mandates to teaching and training further underlie the dif- ferences in the nature of the objectives. The process of drug discovery and development today is vastly different from that which existed prior to 30 to 40 years ago. Failing to recognize this, many today still look back naively to earlier eras to describe examples of how independent academic laboratories can cont- ribute to drug discovery. The changes that have occurred during the last decades, however, have been truly pervasive. Biological and medi- cal sciences have in themselves progressed dramatically, becoming high- ly sophisticated, complex, multidisciplinary, molecularly oriented, and expensive. Regulatory factors have become dominant in drug research, and industrial research laboratories have emerged as new and powerful technological centers. Importantly, the nature of academic research has also undergone significant evolution in form and objectives. The impact of NIH and the grant system has revolutionized approaches, dura- tion (time-range), goals, and execution of projects. There now exists an intense grant-dependency ethic, and the scientific literature and conferences have proliferated almost beyond comprehension. Today, it is not uncommon for academic scientists in biological sciences to speak apologetically of "applied" aspects of their research, as if the purity 376

or excellence of their basic research were to be prostituted by practi- cal objectives. Too frequently in biology is work which is directed to "useful gains" looked upon as being less good for that reason. Yet, scientists of earlier eras are today exalted without explicit apprecia- tion of the fact that many of these were in fact involved in "applied" research in that they continuously sought therapeutically useful goals. Many "academic" drug discoveries of earlier eras would for many reasons today be virtually impossible to reproduce in our academic centers. Most pharmaceutical scientists (pharmacists) and medicinal chemists of earlier times are today concentrated mainly in industrial laboratories. These considerations are raised not for making adverse value judgments, but to acknowledge realities. The notion that science in industry is second-rate or inferior to that of academia is misguided, and those who hold this philosophy retard potential collaboration and information transfer between these sectors. Scientists in industry can be and, on the whole, are as well motivated, creative, imaginative, and devoted to their work as those elsewhere. The view that they are super-technicians or uninspired is totally misplaced. Yet, a degree of elitism and arrogance persists which works to the detriment of the ultimate aims of all scientists in their quest for advancement of science for the benefit of society. In addition to maintaining strong laboratory activities in basic scientific disciplines, industry has to develop highly sophisticated laboratories and groups for toxicology, metabolism, pharmaceutical formulation, scientific information, computing, regulatory affairs, quality control, medical specialties, etc. These activities, which today can consume upwards of 70 percent of the total costs of R&D, are required to bring a compound through the technical and regulatory maze to clinical use. All these activities must always be in close communi- cation so that they function in concert. Suggestions for Promoting Drug Discoveries There are, in my opinion, no single or simple formats for promot- ing or encouraging pharmaceutical discoveries in any disease process or therapeutic field. The entire system of innovation must be studied and considered. As there are multiple defects in our system, so must there be multiple solutions and actions that must be considered. I have listed some areas for discussion, and although considered separately, many are closely interrelated. 1. Need for a constructive public policy supported by sound public opinion Official policy should recognize need for drug dis- covery and integrate this in its objectives for improved medical health- care systems. The economics of drugs versus hospital, surgical or other labor-intensive modalities must be appreciated. The central role of industry in drug discovery should be acknowledged. 377

2. Changes in regulatory requirements and climate Excessive bureaucratic red tape and unnecessary, baseless, and unreasonable requirements for proof of safety and efficacy must be curtailed to accelerate and streamline the drug approval process. Changes in this area could go far to restoring incentives and decreasing the enormous frustrations associated with research. Demonstration of safety and efficacy need not be compromised, but they must be based on sound scien- tific judgments and simple principles of common sense. Rigid regula- tions and FDA administrative "guidelines" and "standards" (such as GMPs, GLPs, Clinical Practices) become petrified and inflexible over time. They not only require establishment of complex and costly systems to deal with them, but worse, they also in themselves discourage creativity and imagination. New tasks and challenges must be accommodated to pre- conceived "standards," and exploratory research (e^£. , in toxicology) is discouraged because of fear of "inspections" that do not recognize this. There are intrinsic incompatibilities between originality and conformity with standard procedures cast in iron. In my opinion, devel- opment and approval processes could be substantially accelerated with- out compromising scientific principles or safety considerations. Over the years, regulatory standards and requirements have become progessively entrenched and accepted, and it is more prudent and expedi- ent for the regulated to adopt and adapt rather than fight losing bat- tles that leave them behind and may incur the wrath of the holders of power. It is amazing but understandable how readily new proposals are accepted or even anticipated by the regulated, regardless of costs and logic involved. A system has consequently been deeply ingrained that will not only be difficult to change but even difficult to understand. Clinical research in particular needs attention, for we are even now heading into ever more restrictive regulatory constraints that severely hamper investigation of new agents in man. Investigators are being literally scared away by, or are finding it professionally repug- nant to deal with, the regulatory framework of clinical drug research. The increasing administrative red-tape, continuous monitoring (by drug firms and regulatory bodies), purposeless "standardization," scrutiny and caution of institutional review committees, patient consent require- ment, responsibilities for future latent drug effects, etc. are becom- ing so overwhelmingly demanding on the academic clinical drug investi- gator that there is a serious risk that the creative, competent clini- cians in this area will be driven away altogether. Another complication is that we have entered an era where, for the sake of "objectivity" and scientific "control" of clinical data, detached principles of statistical theory sometimes virtually dominate performance of clinical studies and trials. This often leads to mad demands that defy simple logic, good medical practice, primary objec- tives, and even ethics. Paradoxically, in our experience, the more novel the drug, the 378

more difficult and tortuous the process, presumably because of the FDA's lack of experience in that area (together with the spirit of extreme caution). The extraordinarily slow process of clinical research increases costs enormously, decreases the crucial feedback needed in the laboratories to introduce further improvements (thus, breaking an essential cycle), and dampens the morale and interest of scientists. We are currently studying some very interesting funda- mental chemotherapeutic leads for leishmaniasis, trypanosomiasis, and malaria. These have arisen from tangential research efforts ("spin- offs"), not from directed, deliberate efforts targeted to these dis- eases. Given the difficulties and expenditures that we are encounter- ing in clinical studies of new drugs intended for important diseases of the United States, we are frankly very discouraged with the poten- tial problems that we would face should these other drugs for tropical diseases merit study in man. For diseases of developing countries a very serious problem is that present law prohibits exporting a drug even for investigational purposes until an IND has been approved in this country. If the dis- ease is very rare or does not exist in this country, it is extremely difficult if not virtually impossible to perform studies abroad under the rigid requirements of the FDA IND system. Certainly, this provi- sion must be re-examined if United States scientists are to be encour- aged to develop drugs for such diseases. Equally significant, the fact that our existing laws prohibit export for sale of any drug unless it is approved for sale in this country does not particularly encourage discovery and development of drugs intended for diseases which only exist in other countries — and to which it is not possible to export. 3. Strengthening of patent systems Because patents and trade- marks are among the few mechanisms by which R&D investments can be pro- tected once a drug reaches the market, they are a crucial component of the viability of industrial innovation. Since the development period is now so long, more than half the patent life may have expired by the time an ordinary new drug reaches the market. This, together with the ease with which many pharmaceutical patents can be copied, challenged, bypassed or infringed without retribution, is causing increasing problems for the innovator even if the drug does reach the market. It must be more widely appreciated that, without the exclu- sivity that patents potentially provide, it is likely that industrial R&D and innovation would virtually cease since in this industry tech- nology is too readily assessed and copied and thus trade secrets give virtually no protection. Anything that can strengthen the patent system, or increase the useful life of drug patents, will add substantial vitality and effort into R&D investments. Insofar as many foreign nations do not honor or enforce patent or trademark laws, further discouragement arises. This becomes particularly serious for drugs intended for primary use in diseases of developing nations. 379

Many potentially useful agents may be known chemical compounds or may otherwise not be subject to strong compound patent coverage. For obvious reasons most pharmaceutical companies will be very reluctant to invest in development of such compounds since there is no protection from copiers. Developing an alternative patent or regulatory policy for such compounds would be encouraging to industry and could result in many new additional medicinal agents. Another change which would sub- stantially help would be the strengthening of use patents by not requir- ing mandatory or compulsory licensing in such cases. An interesting alternative possibility that could to a large extent mitigate the decreasing strength and value of current patent systems would be to assign to the innovator company, through regulatory statutes, a period of marketing exclusivity irrespective of any patent consideration. If this period were to be for 10 to 15 years, it is very likely we would see a very significant stimulus in many areas of drug discovery. It must be remembered that currently many laboratories understandably spend an inordinate amount of effort and time to find the compound which within an active series possesses greatest "patent- ability." If no compound patent protection can be obtained, for exam- ple, because of prior art citations, it is very possible that the entire series would not be pursued further. The assigning of patent rights of discoveries made in universities with public funds to those universities and inventors, with their right to sublicense on an exclusive basis, would help in maximizing the ulti- mate public benefits of those funds. Industry will be understandably reluctant to invest heavily in costly development programs unless there are some assurances (£•£• , through patents) that a reasonable return on that investment can be recovered should the product prove to be useful. A simple fact needs emphasizing: the best potential drug or discovery will not do any good to anyone unless it is developed and manufactured (by an industrial source), and the latter will not be done without some incentive which embraces economic protection. 4. Direct government support and involvement a. Programs in basic research and academic-industrial joint efforts There are several forms of direct federal support that could promote new therapies (drugs and vaccines) for diseases of developing nations, but in my opinion these would be, in themselves, of limited (but definite) value unless truly massive resources (probably unwise) were to be devoted to this effort. Basic research grants in tropical diseases to academic institu- tions, administered through NIH or other governmental agencies through established competitive channels, would certainly be of value. The principal impact would be to focus general attention and interest on these diseases by the American scientific community. This would pro- vide visibility, acceptability, and excitement in this field. It would 380

encourage healthy research competition and general awareness. It would probably be wiser not to encourage "targeted" objectives of drug dis- covery or to fix unreasonable time requirements, but instead provide for long-range fundamental research which the academic community is bet- ter equipped to handle. Funds could probably be relatively modest since the realistic objective would be to kindle and ignite interest rather than solve all the problems. The value would be largely in the form of important intangibles. Workers in other fields, for example, would begin to think about the problems, and discoveries in apparently unrelated fields could eventually be most decisive for the long-term goals of discovering new therapies. An effect of this type of improve- ment of general awareness and credibility of this field of study would be to encourage industrial laboratories to initiate programs in tropi- cal diseases, which would be integrated with their other research acti- vities. It would be less wise and reasonable for industry to ignore activities in areas receiving such important national attention. Such programs should also encourage collaboration with industrial institutions, and joint ventures should be actively promoted. In my opinion, the American research establishment has an enormous, untapped potential for meaningful academic-industrial joint research endeavors in all fields, including tropical diseases. Historically, there has been a void in communication which is most unfortunate because the resources, skills, and expertise are intrinsically complementary and mutually reinforcing. Maximal utilization of our national resources would require that potential collaborations between these sectors, including NIH, be fully developed. What can be done, then, to foster the potential symbiosis of aca demic (university and research institutes) and industrial labora- tories? Clearly, artificial barriers have existed that can be dis- mantled, if the reasons could be fully understood. In my opinion, several correctable factors have prevailed to separate these research communities. Traditionally, industry itself has avoided involvement with programs funded by federal monies since these generally have car- ried patent and ownership provisions which would conflict with poten- tial proprietary rights, and thus restrict economic opportunities and returns on investment. Furthermore, in the past there has been a general philosophy that federal funds should not be expended in any way that could promote the "profitability" of institutions in the pri- vate sector, and that it would be inappropriate to utilize public funds to aid industrial programs or developments. Thus, it has generally been held that discoveries resulting from government-sponsored grants belong best in the "public domain." However, such policies virtually assure that no one will undertake the process of commercialization In order to deliver a product to the consumer! There has been an almost obsessive fear of "misuse" of public funds that may in many cases have actually worked to the detriment of both public and private objectives. Relaxation and enlightened patent policies in this area are in my opinion not only possible but necessary if all our resources are to be 381

used effectively. A deliberate policy to merge certain academic- industrial research activities could be easily promoted while protect- ing potential discoveries, patents, and commitments such as to ensure that the industrial establishment will have adequate incentives for continued development. An imaginative approach in this area, with incentives to both academia and industry, could prove very fruitful. Another factor that has deterred collaborations has been the misunderstanding by many academic centers of the nature and quality of industrial research. The quality and creativity of science in industry are often underestimated, and sometimes their motives are suspect. This is most unfortunate since the level of scientific accomplishment, motivation, and dedication is very high indeed. Industrial scientists work quite independently and in the same overall framework and use the same tools, concepts, and theories of science as do all other scien- tists. Just as the objectives and type of research vary widely among various disciplines within the academic community, so are there similar differences between the orientations of some academic and industrial scientists. Scientists in basic pharmaceutical research laboratories are given considerable freedom and latitude for creative expression, a fact not generally well understood. For substantive, productive col- laborations to materialize, it will be necessary for academic scien- tists to recognize and respect their counterparts in industry because collaborations must be based on equal terms and mutual respect and trust. Greater visibility and recognition of the scientific accomplish- ments by industrial scientists can be of help. If consideration is to be given to funding and establishing special research centers here or abroad which are to be directed to the conduct of studies in tropical diseases, attention should be given to some of the concepts described in this essay that relate to drug research. Care must be taken not to simply create segregated "centers" of tropical disease research. The most productive and cost-effective research in restricted disease states is likely to be conducted under conditions and in an environment which is as broadly based in scienti- fic and medical research as possible. It is likely that establishing relatively isolated "pockets" of strength in a particular tropical dis- ease, although perhaps scientifically sound, are not as likely to yield new drugs. Likewise, it is naive to believe that setting up screening centers to which pharmaceutical companies will send all the compounds off their shelf will be a fruitful exercise. This is simply not the major way drugs are discovered or developed. Meaningful programs in tropical diseases must be closely integrated with other programs which are basic and broadly based in many disciplines. The key to success in this area will be to introduce testing systems and research programs as an integral component of the totality of research activities in a par- ticular setting. This applies to academic and to industrial labora- tories. Similarly, it is most likely that successful innovation in drug discovery for these diseases would arise from tackling these prob- lems in the already sophisticated and advanced laboratories of the most 382

technologically advanced nations instead of setting up de novo centers or institutes in countries where the diseases are prevalent but tech- nology is inadequate. The fastest and most economic approach is not to literally "transfer the technology" across oceans but to challenge those who already have the technology to transfer it internally by the incorporation of additional objectives and programs. In industrial laboratories, new programs and testing systems could be established and tied in with ongoing research relatively easily if the relevant incen- tives and promise for follow-up and economic rewards genuinely existed. b. Federally sponsored clinical evaluation programs This most important area could very substantially stimulate and encourage new discoveries and facilitate development of others that may actually already exist but that linger for lack of proper testing capabilities in man. Although it may seem paradoxical, it is in this area ^i.e«, testing and evaluation in clinical settings), instead of in basic or applied research or discovery, that industry could use help. In addi- tion to difficulties related to availability of proper patient popula- tions for study, especially for diseases of developing nations, many other general problems exist in this area which industry R&D is not as well equipped to handle as are academic and government institutions. Clinical research is becoming increasingly difficult because of FDA, HEW, regulatory and ethical considerations and constraints, and the increasing sophistication of properly controlled clinical trials. Although fundamental knowledge of clinical research, study design, statistical analysis, etc. are well developed in industry, studies must nevertheless be conducted by institutions outside industry by clinical investigators with specialized experience with particular medical problems. Federally-supported clinical research programs conducted through university hospitals or similar institutions would have the advantage for government of having readily identifiable parameters that could be publicly scrutinized and monitored. For example, contract programs to universities or other relevant facilities, administered through federal institutions, such as the NIH or the Armed Forces, could be established. Such programs could be justified to the public because benefits to society would be easier to assess at this stage. Programs in diseases in need of therapies, or rare, "orphan" diseases, including diseases of developing countries, could be specifically selected to spur and encour- age interest and development in these areas. Industry would have lit- tle trouble accepting public support of this nature since admitting this kind (JL.£., clinical) of data into the public arena would not com- promise its patent or other ultimate rights to drugs. In fact, in my opinion industry would welcome help in this area, especially since the data would likely be of high quality and would have public and regula- tory recognition because of having been generated in the public domain and of having been performed under auspices of independent third parties. Some such programs already exist under sponsorship of NIH, such as the extensive cancer testing programs of the National Cancer 383

Institute, the anticonvulsant testing of the Epilepsy Branch of the National Institute of Neurological Diseases and Blindness, and the antiviral testing of the National Institute of Allergy and Infectious Diseases. These and similar programs are already having an important impact in encouraging drug development. Perhaps from these we can learn something that could be applicable to tropical diseases, but some obvious involvement of WHO or other international organizations would be invaluable. c. Improvement of drug and medical care delivery systems in developing nations Whatever can be done to improve general health standards and systems of health care will enchance the likelihood that new (and already available) drugs will reach those who need them. Improvements in these areas will undoubtedly encourage industry to devote greater emphasis to diseases prevalent in these countries.13/ Within the broad subject of prevention of diseases and nutrition- related health factors in developing countries, control of insect vec- tors of diseases by new insecticides, and control of parasitic diseases in food-producing animals deserve some mention. The veterinary research programs of major American pharmaceutical companies recognize the close interplay between drug development for both medical and veterinary indications. At this time, industrial research is providing one of the few effective bases for coordinated efforts to develop veter- inary medicinals and insecticides. d. Direct subsidies to industry by government In my opinion, such mechanisms of sponsoring drug research programs would be of little value and would not likely be acceptable to many in industry. Industry already has the potential expertise and organization for such research programs. The reason it may not be pursuing these areas as vigorously as it could is not so much lack of additional funds required for those specific programs, as the increasing disincentives to basic research programs in general and the perceived, economically unproductive nature of drugs in this particular area. Equally important, research activi- ties in all therapeutic fields are so interrelated in industrial labs that it would not be possible to clearly distinguish exactly which activities had and which had not been supported by federal funds. Since this could challenge the propriety or ownership of unexpected dis- coveries and "spin-offs" in other areas, I suspect that most in indus- try would not use such funds unless they were in very well-defined areas (e_.^., vaccines). Overall, this form of support would generally be short-sighted unless it were coupled to other major changes in atti- tudes and legislation affecting the industry. It must be remembered that once a candidate drug has been "discov- ered" in the laboratory, the actual costs incurred in having made that discovery are very likely to be minuscule compared to subsequent costs of developing that drug for introduction to the market. However, the real costs of the research for that particular discovery are actually 384

the costs of maintaining the entire research programs of that company, including the many things that fail. There are many "discoveries" of chemicals with potentially useful properties which do not survive the rigorous safety and efficacy evaluations required before introduction into medical practice. Even considering only those drugs that are introduced for formal study in man (as IND), less than 10 percent are ultimately submitted (as NBA) for marketing approval.ll/ e. Special tax incentives for industrial R&D expenditures Special incentives would certainly be achieved by allowing additional or special tax benefits, allowing for capitalization of such research investments to those pharmaceutical companies who invest a substantial proportion of their sales in R&D. 5. Identification of new lines research In my opinion, innumer- able, exciting opportunities exist for applying our current body of fundamental knowledge and technology to diseases of developing countries. As we have already heard, advances in concepts and technology provide sound foundations for rational attacks on the basic causes of these diseases. The tools of molecular biology, protein chemistry, molecular modeling, and comparative biochemistry are now particularly powerful for understanding and possibly controlling diseases caused by parasitic, rickettsial or bacterial agents, or those mediated by insect vectors. Several parasites can now be cultivated in tissue culture under control- led conditions, providing new opportunities for biochemical studies and for characterization. Identifying specifics of what can be done, or particular strategies in specific disease states, is not the real prob- lem since these are well known to specialists in the laboratories. The scientific details of programs will be forthcoming provided the overall climate is such that it will allow and encourage these scientific oppor- tunities to express themselves through the best suited, natural, and economic channels. Summary The optimal means for encouraging new drug discovery for diseases of developing nations are primarily to deliberately acknowledge the desirability of drug innovation in general. If trends which now dis- courage or prevent drug discovery in general can be reversed, I am con- fident that today's advanced state of science and technology will in the future provide society with new, innovative therapies for many diseases, including those occurring in developing nations. In order to begin to reverse the currently reduced degree of commitment to drug discovery, and the prospect of an ever dwindling rate of new drug intro- ductions, it is essential to understand the complex forces, mechanisms, and sources involved in drug innovation and discovery. The pivotal role of the research-based pharmaceutical industry in this process must be clearly acknowledged and fostered. 385

REFERENCES AND NOTES 1. In this essay "drug discovery" is used primarily to denote the process of identifying candidate drugs while the closely related process of "drug evaluation" involves several sequential stages of drug safety evaluation and clinical efficacy trials. As will be discussed, for important medical diseases which are virtually nonexistent in this country (£•£. , Chagas' disease, filariasis, malaria), American industry has sound resources for "discovery" but will need cooperation with developing countries and government agencies for the phases of drug evaluation. 2. A notable exception may be found in the National Cancer Institute's cancer testing (and screening) program, which, however, has been extremely expensive and is directed not so much to a true drug discovery process as to testing of compounds synthesized by other sources. 3. This has occurred largely through funding by the government, which ironically has been inhibiting its use. 4. Schwartzman D: Innovation in the Pharmaceutical Industry. Baltimore, Johns Hopkins University Press, 1976 5. Regulation, Economics, and Pharmaceutical Innovation, Proceedings of the Second Seminar on Pharmaceutical Public Policy Issues. Edited by JD Cooper. Washington, DC, The American University, 1976 6. De Haen P: Compilation of New Drugs, 1940 through 1975. Pharmacy Times, 42:40-74, 1976 7. Participation of industry in development of drugs for tropical diseases is not a new or recent commitment. It is of historical interest that establishment in 1901 of the Wellcome Tropical Research Laboratories at the Medical College of Khartoum in the Egyptian-Sudan actually preceded that of the research center in London which was later named the Wellcome Laboratories of Tropical Medicine. There has been a continuing allocation of resources within the Wellcome Research Laboratories to research programs from which candidate drugs are identified for evaluation against parasitic, viral, and bacterial infections. 8. A just-released two-year study sponsored by the National Science Foundation 10/ has concluded that there has been a real and con- tinual decrease in industrial basic research which should be of serious national concern. This trend was seen as "another 386

manifestation of the existing poor climate for private enterprise that is observable in the unwillingness of corporate management to commit resources to the future — i«£« , take risks." Among sever- al factors responsible, the effect of the regulatory environment was described as "pernicious," and in this context "the problem is that these ruinous outcomes on fundamental research activity are never direct — they impact on the way an organization sets R&D priorities." The study concludes that "many steps must be taken by the public sector to improve the climate for risk taking in the private sector." 9. The long history of efforts and contributions in tropical diseases by Burroughs Wellcome and the Wellcome Foundation has seen some notable accomplishments: successful development of yellow fever vaccine (1928); introduction of sulfones for treatment of leprosy (1938); discovery of pyrimethamine for malaria (1952); and use of methisazone as the first antiviral agent effective against small- pox (1963). A continuing effort in chemotherapy and on development of vaccines for parasitic diseases are part of the continuing research effort. 10. Support of Basic Research by Industry. Report prepared for the National Science Foundation by a grant to Industrial Research Insti- tute Research Corp and Rensselaer Polytechnic Institute, 1978 11. Wardell WM, Hassar M, Anavekar SN, Lasagna L: The Rate of Develop- ment of New Drugs in the United States, 1963 through 1975, Clin Pharmacol Ther 24:133-145, 1978 12. For example, of the 79 NCEs introduced between 1962 and 1968, the majority achieved less than $2 million in sales in 1972, and 33 (42%) were under $1 million._14_/ Only eight achieved a level of $16 million, a figure computed to be necessary k_l to achieve a return on investment of 10 percent after taxes. 13. It is generally recognized that many developing nations cannot "afford" available drugs, and the solution to this problem has often been directed to the pharmaceutical industry. Industry is asked to make these drugs available free or at cost, for humani- tarian reasons. This philosophy unfairly ignores the basic nature, objectives, and functioning of this industry. Furthermore, and more importantly, it apparently does not recognize that procure- ment and delivery of health care and drugs is basically the respon- sibility of the State, and that subsidization must therefore derive from public bodies and not from private firms. 387

14. De Haen P: New Product Survey and Nonproprietary Name Index; and U.S. Pharmaceutical Market, Drug Stores and Hospitals, IMS America, Ltd. (Note: new esters and salts are included in sales of the new single chemical entity.) 388

DISCUSSION: THE SCIENTIFIC BASE: OPPORTUNITIES FOR RESEARCH DR. LEDERBERG invited Dr. Lucas to comment on Dr. Cuatrecasas• advocacy of centering basic drug development research in places where the requisite expertise exists, instead of centering it in countries where the diseases occur. DR. LUCAS, agreeing generally with Dr. Cuatrecasas, emphasized the necessity also to strengthen clinical research centers in developing countries, where clinical evaluation of new drug compounds must be per- formed. He acknowledged that the majority of new drugs would come from industrialized countries, where screening and testing of compounds in the initial stages can be done far more efficiently. DR. CLUFF was interested in the imperative to find new avenues of communication between the scientific community and the industrial world. Referring to Dr. Cuatrecasas' pleas for improved relationships between the academic and industrial research communities, he requested the panel to address the question of how better communication can be fostered. DR. JOHN URQUHART, drawing on his personal experience in both worlds, compared the narrower focus or perspective in the academic world, which digs deeper into the scientific base of problems, with the broader range and shallower depth in industry's search for practical applications of new discoveries. DR. LEDERBERG deplored the fact that graduate students in the basic biological sciences have very little exposure to industrial per- spectives, or understanding of the drug development process. He agreed that the overwhelming majority of practical new drug discoveries have come from industry, and sight should not be lost of the relative order of rule between academic and industrial contributions. DR. de MAAR said that WHO had faced the same lack of communication between industry and academia in the early days of the Tropical Disease Research Program. A helpful solution was development of the Scientific Working Group concept, which brought together for continuing dialogue scientists from industry and from academia, and from developed and developing countries. This relationship between the industrial and academic communities has proved mutually advantageous and complementary. DR. SEYMOUR COHEN remarked that the organization of this Conference is an indication of the deteriorating position of academia as compared with industry and government. He noted that the first two days of the 389

Conference were devoted to policy questions without considering the content of scientific issues that should have been examined first or as a basis for developing sound policy. Thus, there has been no overt recognition of the fact that progress in infectious diseases research generally will heighten the probability for also making useful discover- ies or finding solutions to the problems of tropical diseases. One can- not consider tropical disease research separately from infectious dis- eases generally, since the scientific aspects of these two fields are intrinsically linked. DR. ALVAN FEINSTEIN pointed out that the attitudes of his academic colleagues also serve to block communications with industry. He charac- terized the prevailing feeling among academics that they were "noble knights on horseback" holding the "evil dragons of industry" at bay. The profit-making motives that influence industrial decisions are analo- gous to the power and influence motives which pervade academic power struggles. DR. LEDERBERG summarized by noting that: industrial and academic scientists belong to the same species, Homo sapiens. 390