156 ANTIBACTERIAL SUBSTANCES FOR TREATMENT OF INFECTIONS 



to a fairly constant pattern, the ratio is a valuable index of therapeutic efficiency. 

 For example, using rats suffering from an experimental trypanosomiasis, the 

 ratio of the minimal lethal dose to the minimal curing dose is unity for arsenic 

 acid and 4 for atoxyl. Arsenic acid is clearly useless, and atoxyl not very satis- 

 factory. Under the same conditions, arsphenamine has a ratio of 37. 



Although a great deal of precise chemical investigation accompanied the search 

 for chemotherapeutic agents, we have to-day little knowledge that will enable us 

 to predict what type of substance is likely to prove effective against a given parasite. 

 The successful discoveries have been largely the result of trial and error. Widely 

 different substances prove to be similar in their efficacy on one type of infection, 

 though once a type of substance has proved efficacious it is usually possible to 

 relate its molecular structure to its chemotherapeutic activity. For example, 

 trivalent arsenic attached to a benzene ring is most effective against trypanosomes 

 when there is an amino group in the para position, and against spirochaetes when 

 there is a hydroxyl group in the para position and an amino group in the ortJio 

 position. Thus, arsphenamine has the formula 



Hg N NHg 



ho/ \As=As/ \oH 



Another general point of importance is that the drug does not necessarily 



act in the form in which it is administered. Atoxyl, H2N<' yAsOg HNa, for 



example, is a pentavalent arsenical, but is reduced to the active trivalent form in 

 the body. 



It is supposed that chemotherapeutic drugs act either by weakening the parasite 

 so that it is easily eliminated by the defence mechanisms of the tissues, or by 

 killing the parasite outright. Experience with experimental trypanosomiasis 

 shows that, if small doses of the drug are given, the " weakening " process may 

 not only be ineffective, but that subsequent generations of the parasite may 

 develop immunity to the drug and become " drug-fast." 



CHEMOTHERAPEUTIC AGENTS ACTIVE AGAINST BACTERIA 



The development of chemotherapy in spirochaetal and protozoal infections 

 in the early years of the century had no parallel in bacterial infections. In 1911 

 Morgenroth and Levy (1911) were able in a certain proportion of cases to protect 

 mice against pneumococcal infection by ethyl dihydrocupreine, but apart from 

 this there is little to note until 1935, when Domagk (1935a, b) reported that the 

 compound sulphonamido-crysoidin (Prontosil) would cure streptococcal infection 

 in mice. The curing dose was between one-tenth and one-fiftieth of the tolerated 

 dose for the acute infection that followed the intraperitoneal injection of ten 

 lethal doses of Str. pyogenes, and between one-hundredth and one five-hundredth 

 for a less acute infection. The interval between infection and the start of suc- 

 cessful therapy was in some cases as long as three days. Domagk's observations 

 were quickly confirmed by French and British workers (see Levaditi and Vaisraan 

 1935a, b, Nitti and Bovet 1935, Buttle 1935, Colebrook and Kenny 1936). Sulphon- 

 amido-crysoidin, however, was active only in the animal body, and had no effect 

 on streptococci in vitro. Trefouel and his colleagues (1935) supposed that the 

 in vivo action was due to a breakdown product of the drug, and showed that the 



