I39

THE MOST RECENT ANTIBIOTIC AGENTSBy GEORGE BROWNLEE, PH.D., D.Sc.

Department of Pharmacology, University of London, King's College

The forfeit which medicine pays for the fruitfuldiscovery of antibacterial chemotherapy is theembarrassment of riches. The corrective, shouldit prove necessary, is provided by recalling to mindthe state of affairs which existed before Domagk's(1935) discovery of the chemotherapeutic activityof prontosil rubrum against infections due toStreptococcus pyogenes. So little success had re-sulted from attempts to apply the classical chemo-therapy of protozoa and spirochaetes to bacteriathat in many quarters failure was presumed to beinevitable. Indeed, the slow development whichhad characterized the use of mepacrine, claimed byKikuth in I932 to be a useful drug for the treat-ment of malaria, might well have been duplicatedwith prontosil rubrum but for the careful work ofColebrook and Kenny (I935) in puerperal sepsis.The significance of this discovery could not beunderestimated, and an explosive burst of energyresulted. Thus, by I940, with the discovery ofthe substituted sulphonamides, pneumococci hadbeen brought within the scope of action and therange widened to include staphylococci. Thestage was now set.

Interest in mutual antagonism between micro-organisms is as old as Pasteur's fundamentalstudies, but the many researches undertaken todayin this field derived their impetus from thedemonstration of the chemotherapeutic nature ofpenicillin by Florey and his collaborators (Chain,Florey, Gardner, Heatley, Jennings, Orr-Ewingsand Sanders, 1940). These observations, in turn,started with the inhibitory action of the mouldcontaminant observed by Fleming (I929). Thesingular freedom from toxicity of even grosslyimpure samples, when taken together with itscatholic range of antibacterial and antispirochaetalactivity, distinguishes it as the most remarkablechemptherapeutic agent ever discovered.The methods which were developed for the

selection of high-yielding strains of penicilliumnotatum, and those which emerged for the extrac-tion and purification of the antibiotic provided apattern which was to be followed and duplicatedwith bewildering. frequency. The effect was tobroaden the range of pathogens brought within thescope of chemotherapy and to increase theefficiency of known examples.

The result has not been an unalloyed delight tothe physician, since the claims of alternativechemotherapeutic antibiotics are inevitably pressedbefore adequate therapeutic comparison has yetestablished the final place of their predecessors.Nevertheless the discipline of adequate thera-peutic comparison remains the minimal standardby which a final assessment may be made. In thisway the established position of sulphonamides,penicillin and streptomycin is subject to constantreview; and in the same way must the challengeof aureomycin, chloramphenicol and terramycinbe proved. Thus it remains the duty of thephysician to weigh and balance the claims of newdiscoveries against those of well-tried agents, as itis also the duty of the pharmacologist to advancethe claims of others. The minimal pharmaco-logical ?equirements which are the essential pre-requisite to clinical trial are now well understood.At this stage of our review it is pertinent to

enquire whether the experiences which haveguided the choice of currently accepted antibioticsfurnish any clue to current and future claimants.Let us turn first to this aspect.Factors Affecting the Choice of New Anti-biotic Agents

I. The antibiotic affects pathogens higherto in-susceptible to chemotherapy. A new antibioticwith an adequate therapeutic index* effective indiseases caused by viruses, in infections caused bysome gram-negative rods, such as proteus, in someprotozoal infections, such as hepatic amoebiasis,will be pressed into future service.

2. The antibiotic affects pathogens susceptibleto existing chemotherapy but is less toxic. Anequally effective agent which proved less toxicthan streptomycin to the eighth nerve would be ex-pected to displace its use in tuberculosis. Similarlyan equally effective agent with a wider margin ofsafety would be expected to displace polymyxin

Average Lethal Dose*Therapeutic Index= -Minimal Effective DoseLLDSM.E.D.

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from

140 POSTGRADUATE MEDICAL JOURNAL March 1952

B or E in meningitis of Haemophylus influenzae orPseudomonas pyocyanea origin.

3. The antibiotic is equally effective yet is lessprone to facilitate emergence of resistant strains.The rapidity with which organisms develop re-sistance to streptomycin in clinical use constitutesits greatest disadvantage, and may be expected tocondition its replacement.

Certain members of the gas-producing anaerobesand the causal organism of tuberculosis are notableomissions from paragraph I. It is probably not tobe expected that the discovery of an intrinsicallymore efficient antibiotic would alter the presentday concepts of management of these infections.It seems that the pathogenesis of the diseasedictates its course and the elimination of thecausal organism by antibacterial action may notremove the immediately toxic factor. If we turnto the experiences given by tests of comparableefficiency of antibiotics and other chemotherapeuticagents in controlled animal experiments on avariety of genera, the underlying reasons begin toappear.

In their more usual forms the tests are designedstrictly as biological standardizations; that is tosay, as an assay in which one drug, posing as astandard, is compared with another. The end-point is a convenient biochemical reaction causedby the parasite. They are therefore analogous tothe standardization of insulin or of diphtheriatoxin. They are not analogous to the clinicaltreatment of diabetes or diphtheria and, of course,they resemble only faintly the clinical chemo-therapy of tuberculosis. Nevertheless, tests of thiskind supply much additional evidence; it is onlywhen we misuse this information that we are ledastray. This often happens.Animal TestsThe amount of information to be extracted from

an animal test is restricted by its design. A studyof the development of chemotherapy reveals thereasons for modelling tests upon those whichproved successful in the development, first ofprotozoal and then of gram-positive and, later,gram-negative bacterial chemotherapy. Tests ofthis kind, which might well be described asprophylactic tests, are usually conducted in thefollowing way. Groups of animals, often mice,are injected parenterally, usually intraperitoneally,with a number of lethal doses of the pathogen, andare divided into groups. The treated groups aregiven the drug, usually immediately and con-tinued at short intervals for a shorter or longertime, and the assessment of worth is usually basedon survival time or a similar parameter.

Provided the principles of biological assay arerespected, and notably the inclusion of a control

group, infected but untreated with drug, and thesimultaneous inclusion of an active substance asa standard of comparison, the method may bereliable and capable of statistical appreciation.May we remind ourselves that simple tests of

this kind facilitated the discovery of prontosilrubrum, sulphanilamide and the substituted sul-phonamides, streptomycin and the newer anti-biotics, and remain today the basis of screeningtests for gram-positive and some gram-negativeorganisms. Indeed, in one form or another thissimple test was used as the starting place for theadvancement of most chemotherapeutic substancesknown to us.The test organisms included are very variable

among both gram.positive and gram-negative or-ganisrms. It should be noted that the design of thetest requires an overwhelming infection which issuppressed during the course of the drug ad-ministration, and for a short time thereafter; forif the survivors are kept they invariably succumbto a delayed bacteraemia.

Experiments in which Toxins are ProducedIn contrast we must now look at experiments

made with micro-organisms which produce one ormore diffusible toxins. Should the infecting or-ganisms be drawn from a heterogeneous groupwhich produce diffusible toxins, e.g. Cl. tetani, Cl.botulinum, Cl. Welchii or N. diphtheriae, allorganisms susceptible to sulphonamides, then thetest is likely to fail, even on a prophylactic level,unless special precautions are taken. Thus, asatisfactory comparison of the ' gas-gangrene 'clostridia may be made, provided the infectingorganisms are first washed free from toxins. It isinteresting and significant that the spores liedormant in the apparently healthy tissue of animalswhich have recovered from the primary lesion andmay be reactivated at a later period by the in-jection of necrotising agents such as sterile quartzor calcium chloride solution.At least two biochemical lesions are involved;

the first, the bacteriaemia of numbers of or-ganisms or the non-diffusible somatic toxin, andthe second, the specific biochemical lesion of thetoxin itself. There is usually more than one of thelatter. The animal test here is a faithful mirror ofclinical experience with chemotherapy againstthese causal pathogens.Experiments in which Numerous Bio-chemical Lesions are Produced. Acid-FactOrganisms

Let us now turn to tests in which acid-factbacteria are used and which produce a number ofpharmacologically active substances. At once amajor difference is met. Instead of an acute ex-

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from

March i952 BROWNLEE: The Most Recent Antibiotic Agents 141

periment of a few days' duration based on anoverwhelming infection, we must resort to achronic infection of some weeks' duration basedupon an admittedly large number of organisms.

In a simple test with mice it is usually essentialto inject a large number of viable organisms intra-venously to ensure the death of all animals withina group in the short period of time of 30 days.With guinea-pigs, injection of a large number oforganisms is usually made subcutaneously, and thetest is of a number of weeks' duration. More re-cently tests in dogs and monkeys infected intra-tracheally or intranasally have been described byFrancis, Spinks and Stewart (1950). The lattertests have arisen from an appreciation of thepeculiar pathogenesis of the disease. However, alltests illustrate the same principle, namely, thegenesis of a number of biochemical lesions as-sociated with'the pathology of the disease, andwhich modify its course. Thus, instead of abacteraemia we have a characteristic lesion arisingas a direct result of multiplication of the parasite-the tubercle. Associated are a number of bio-chemical lesions, discussed elsewhere (Brownlee,1948), which may be listed.

The Biochemical Lesions which Follow In-vasion by a Tubercle Bacillus

I. Multiplication of the parasite.2. Chemiotaxis of monocytes.3. Inhibition of proteases.4. Hypersensitivity effect.5. Skin allergy.6. Acauired resistance.Specific antimicrobial chemotherapy is as-

sociated only with the first, and even this isquickly modified to create the limitations as-sociated with the use of our currently usedchemotherapeutic agents. We may now acceptthe hypothesis that some disease processes, e.g.,those caused by clostridia and tuberculosis, areunlikely to be dramatically modified beyond ourpresent understanding by the discovery of a moreefficient antibiotic. There are, however, someadditional factors which pass comment on thisthesis.

The Factors which Influence the Effect ofChemotherapeutic Agents on a BacterialPopulation.These may be listed as:

I. The efficiency of the agent.2. The effect of numbers of organisms.3. The rate of emergence of resistant strains.

Whether factors i and 2 are related depends onthe mode of action of the drug. For example, itis agreed (Henry and Hobby, 1949), that strepto-mycin is predominantly bacteriostatic, although it

may be shown to be bactericidal under certain con-:ditions. Unlike penicillin, the activity of strepto-mycin varies with the concentration, the size ofthe inoculum, the medium and the pH. Strepto-mycin is most effective against dividing cells, butmay have some effect against resting organisms.A present-day limitation to the unrestricted use

of streptomycin in chemotherapy is the frequencyand rapidity with which resistant strains emergefrom initially sensitive strains. This is found'alsowith test-tube experiments. In contrast, theemergence of resistant strains to penicillin is low.The streptomycin resistance ranges from cells onlyslightly more resistant than the original to absoluteresistance (Demerec, 1945); with penicillin therange is narrow.

Test tube experiments show two patterns ofdevelopment of resistance. Resistance developsin one pattern, characteristic of penicillin, only ina slow, step-wise manner during serial transfer inprogressive concentrations of the antibiotic. Inthe second pattern, characteristic of streptomycin,a high degree of resistance may develop after com-paratively few exposures or possibly after even asingle exposure to the antibiotic (Miller andBohnhoff, I946).Mechanism of Action of Emergence ofResistanceThe statistical method of Lauria and Delbriick

(I943) has provided acceptable evidence for themutation theory. In brief, it is believed that in anylarge population of bacteria there exists, albeit inrelative small numbers, mutants capable of sur-viving and growing in otherwise bactericidal con-ditions. In criticism one must acknowledgevariables other than mutation to be included in thestatistical population. On the other hand, thosewho are critical of the adaptation theory, point tothe existance of strains resistant to concentrationsof antibiotics greater than that in which they havebeen trained. (See, for example, the experimentsof Silver and Kempe, 1947; Demerec, I945;Linz, 1948.) A recent discussion by Barer (195')is of interest, and for the present the subject muststill be considered to be under discussion.

The Logical Use of Combined TherapyThe inability of micro-organisms to grow in a

given concentration of an antibiotic involves theselection (elimination) of those cells unable tosurvive in that concentration. The survivors arethe progenitors of the ultimate resistant popula-tion. Other antibiotics, antibacterial by reason ofa different mode of action, may be logically used toeliminate the susceptible fraction of the emergentpopulation. In theory, as many agents as there aremutants may be used. In practice, two or three

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from

142 POSTGRADUATE MEDICAL JOURNAL March 1952

drugs are used simultaneously. The practicalproblem resolves itself into bridging the desiredperiod of treatment without emergence of a re-sistant strain to any one drug.

Combinations must, however, be justified byresults. Diminution in efficiency may be met whentwo agents are combined. This is possibly true ofthe combination of aureomycin with penicillin,which in laboratory experiments (Gunnison,Coleman and Jawitz, 1950) is less effective thanwhen used singly. It appears that the bacterio-static action of aureomycin inhibits the metabolicprocesses of growth with which penicillin inter-feres. From our knowledge of the similarity ofaction of aureomycin, chloramphenicol and terra-mycin, the latter two drugs might be expected tobehave in a similar way.Classification of AntibioticsFrom the basis of this review we are now in a

position to consider four groups of antibioticswhich fill the present scene.

Penicillin has been called an ideal chemo-therapeutic agent. It is a weak acid whose saltsare soluble, are rapidly excreted and are withouttoxicity. It has been studied most and we maysoon expect to know something of its mode ofaction (Brownlee, 1951). Mycomycin, a new anti-mycobacterial substance, also a weak acid, is apotential candidate for this group.A second but basic group, also derived like

penicillin from fungi, includes streptomycin (andstreptothricin), neomycin and viomycin. Theseantibiotics are stable, soluble in water and in-soluble in otganic solvents. These substances aretoxic but useful. A beginning has been made withmode of action studies of streptomycin.A third group comprises those extracted from

bacteria. They are basic polypeptides of highmolecular weight. Even when purified by exactingand meticulous physico-chemical tricks, they areusually mixtures of closely related substances.They are thermostable, diffuse slowly, adsorbreadily and are absorbed and excreted slowly.Some are quite toxic; some, like polymyxin Band E are not. Included are the polymyxins,gramicidin, subtilin, circulin, bacitracin andlicheniformin. Nothing is known of their mecha-nism of action.

In a final group may be included antibioticswhich show antirickettsial activity. Aureomycinand terramycin may prove to be similar chemically.Both are amphoteric substances reacting with acidsand bases to form soluble salts. Chloramphenicolis a neutral substance. The mode of action of thisgroup is unknown.The MOst Recent Antibiotic AgentsThe rapid rate of excretion of penicillin

necessitates frequent injections, and alternativedevices have been sought to maintain tissue con-centrations. These include (a) the use of verylarge doses, (b) the use of mechanisms to delayabsorption, and (c) the use of drugs to minimizekidney excretion. Mechanism.(c) is related to thefact that 80 per cent. of urinary penicillin is ex-creted by the tubules (Rake and Richardson, 1946).By simultaneously giving large doses of drugswhich are also excreted by the tubules, penicillinexcretion is delayed by competition. Examples ofthis largely theoretical use are diodone (Rammel-kamp and Bradley, 1943), p-aminohippuric acid(Beyer et al., I947), and sodium benzoate. (Spauld-ing et al., 1947). Beyer et al. (i947) ascribed ad-vantages to 'Caronamide' (carboxyphenylmeth-anesulphonanilide) which reversibly blocks ex-cretion of penicillin, it is believed by enzymic in-hibition, although itself excreted. slowly onlythrough the glomeruli. The most useful exampleof mechanism (b) proved to be the use of procainepenicillin, which retains sufficient of its localanaesthetic properties to be painless on injection.It has the disadvantage that it inhibits the actionof sulphonamide drugs by reason of its procainecontent. For a newcomer, dibenzylethylenedi-amine penicillin (Elias, Price and Merrion, I951),is claimed the advantages of better duration-concentration time than procaine penicillin, andthat repeated doses gives cumulative effects. Ithas the same anaesthetic potency but is slightlymore toxic. For an ester of penicillin marketedunder the trade name of' Estopen' is claimed theadvantage of excretion into the sputum in highconcentrations. Preliminary experiences are saidto be encouraging.Streptomycin and DihydrostreptomycinDanger of temporary or permanent damage to

the eighth cranial nerve, resulting in vestibulardysfunction, is in direct ratio to the dose andduration of streptomycin treatment. The hopethat this danger might be minimized by the use ofdihydrostreptomycin has not been sustained. Itfollows from the earlier discussion that we mayconfidently expect the development of additionalantibiotics with advantages over streptomycin andits reduction product. These would includebactericidal rather than bacteriostatic action, in-susceptibility to the effects of numbers of or-ganisms, and a minimal rate of emergence ofresistant strains..

Neomycin. Neomycin is a product of Strepto-myces fradiae. It has antitubercular action in thetest tube, including streptomycin resistant strains(Waksman and Lechevalier, I949). It is. also,chemotherapeutic in the tubercle bacilli infectedanimal (Carr, Pfuetze, Brown, Douglas and

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from

March 1952 BROWNLEE: The Most Recent Antibiotic Agents 143

Karlson, 195 ). Serious toxic effects, the principalof which are renal damage and deafness, which isprogressive after administration is stopped, areobserved. Unless these prove to be due to im-purities which can be eliminated the antibiotic isunlikely to survive clinical trial.

Viomycin. Isolated from Streptomyces floridaeor puniceus in two different laboratories in I950,it is antibacterial in vitro against Myco. tuberculosiswith little or no other antibacterial activity. Inthe experimental animal its chemotherapeuticefficiency is little different from that of strepto-mycin. It causes renal damage and vestibulardamage, and also partial deafness. Changes inelectrolyte balance with a fall in plasma potassium,calcium, phosphorus and chlorides and an increasein CO2 combining power of the blood have beenrecorded. It has been pressed into serviceagainst streptomycin-resistant tubercle bacilli(Finlay et al., 1951).

The polymyxins. The name polymyxin appliesnot to one substance but to a group of five closelyrelated polypeptide antibiotics derived from fivedifferent strains of Bacillus polymtyxa. PolymyxinA was originally known as ' Aerosporin ' and wasshown to be nephrotoxic in animals by Brownleeand Bushby (I948). Polymyxin D is the originalpolymyxin described by Stansley et al. (1947), andwas shown to be similarly nephrotoxic in animals(Bryer et al., 949; Brownlee et al., I949; Schoen-bach et al., I949). A third nephrotoxic memberof the series, polymyxin C, has also been reported(Brownlee et al., I952). Although the renaldamage caused by these polymyxins was transitoryand limited to the convoluted tubules it wassufficient to restrict their use, and a fuller clinicalassessment could not be made until this renaltoxicity had been overcome. Polymyxin B was re-ported by Brownlee et al. (I949) to be virtuallyfree from this toxicity, but to cause pyrexia, neuro-toxic signs and severe reactions at the site of in-jection in animals. Polymyxin E, the most recentof this group, does not cause pyrexia or local re-actions in animals and is almost free from the renaltoxicity displayed by polymyxins A, D and C(Brownlee et al., I952). Apart from these pharma-cological differences the polymyxins are similar.They have almost identical antibacterial activitiesin vitro, being specifically active against the gram-negative bacilli, and all containing approximatelyIo,ooo polymyxin A units per mg. Their action isbactericidal and it is only with difficulty that re-sistant strains can be made to emerge from re-peated sub-culture. The minimal inhibitory con-centration depends on the number of organismspresent. Experimentally the polymyxins havesome advantages over streptomycin. They appearto be the most effective agents known to us against

Haemophylus pertussis, H. influenzae and Pseudo-monas pyocyanea. Although streptomycin is alsobactericidal, its intrinsic activity is less than thatof the polymyxins (Alexander et al., I950);streptomycin is active in concentrations of io0g.per ml., whereas polymyxin is active at i to Io (g.per ml. Brownlee and Bushby (1948) showedthat the in vitro strain CN 258 of H. influenzae wasinhibited by o.oz ig. per ml. of polymyxin A com-pared with i (g. per ml. of streptomycin, and inmice infections Ioo pIg. of polymyxin A gave thesame protection, as 1,250 vg. of streptomycin.Streptomycin readily permits resistant organismsto develop, but with the polymyxins resistantstrains are produced in vitro only with the greatestdifficulty. In human cases of H. influenzae in-fection treated with polymyxin there has not beenany tendency for resistant strains to develop(Alexander et al., I949).

Alexander and her colleagues (I949) comparedthe in vitro activities of aureomycin and chloram-phenicol with that of polymyxin B. They foundthe minimal effective concentrations required toprevent growth of H. influenzae in vitro to be:Streptomycin, I to 3 ,ug. per ml.; polymyxin B,0.3 tzg. per ml.; chloramphenicol, i v±g. per ml.;and aureomycin, 0.5 to I.0 Itg. per mi.

Chandler and Hodes (I950) compared thetherapeutic effect of streptomycin, dihydrostrepto-mycin, polymixin and aureomycin in H. influenzaetype B infections in mice. At high doses there waslittle to choose between the antibiotics, but at lowdoses polymyxin B was much the best.

Swift and Bushby (1951) successfully treatedeight children intrathecally and intramuscularlywith polymyxin B (one) and polymyxin E (seven)suffering from H. influenzae meningitis. Brakley(1950) reported the successful treatment of asevere case of H. influenzae meningitis with poly-myxin B, and Hayes and Yow (1950) have de-monstrated its value in meningitis caused by Ps.pyocyanea. In view of the interest being takenin the antivirus potedtial of antibiotics, the anti-phage activity of polymyxin B is of interest. Of45 substances of known antibacterial activity whichwere tested against 60 races of bacteriophage,using paper disc methods, only eight showed anti-phage activity, and only polymyxin B showed goodactivity. Against one it proved active at a con-centration of io mg. per ml. (Hall, Kavanagh andAsheshov, 1951). Subtilin, which was included inthis test, appears to have comparable phage andvirus activity, since it is claimed to be capable ofinactivating influenza A virus (Salle, 1950).Aureomycin, Chloramphenicol and Terra-mycin

These three antibiotics have established for

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from

44 POSTGRADUATE MEDICAL JOURNAL March 1952

themselves a place in medicine not only for theircatholic antibacterial spectrum but by breakingnew ground with antivirus and anti-rickettsialactivity. They are essentially similar and includein their fields of action gram-negative and gram-positive pathogens. They therefore find a naturaloutlet against penicillin-resistant and strepto-mycin-resistant organisms. Current research isconcerned with establishing the particular sphereof each, and we will concern ourselves only withspecial aspects.

The simultaneous increase in resistance of bacteria.The simultaneous and parallel increase in re-sistance of organisms to aureomycin, terramycinand chloramphenicol on exposure to each anti-biotic in vitro and in vivo has been commented onby different groups of workers (Herrell, Heilmanand Wellman, I950). Terramycin appears tohave the advantage of a pattern of microbial re-sistance characteristic of penicillin in contrast tothat of streptomycin (Hobby, Lenert, Donikianand Pikula, 195I). Terramycin appears to beestablishing a special use in syphilis. Unlikeaureomycin and chloramphenicol, it resembles theactivity of penicillin in healing the scrotalsyphilomas caused by Treponema pallidum inrabbits (Levaditi and Vaisman, 195I). Terra-mycin has been described by Loughlin andJoseph (I95i) as the most effective antibiotic yetstudied in yaws and tropical ulcer. Aureomycinhad been found of value in the treatment ofamoebic dysentery. Clinical improvement with85 per cent. cures, with disappearance of amoebae,being reported by Siguier and Choubrac (I95I).Hepatic amoebiasis was unaffected. A sober andvaluable assessment by Ley et al. (I951) observedclinical control in x2 of 12 patients with re-

appearance of the amoeba in three of eight patientswho were followed up. The same authors drawattention to similar results observed by Armstrong,Wilmot and Elsdon-Dow (1950), who are in apeculiarly favourable position to judge by reasonof their past experiences and trials.

Chloramphenicol. This antibiotic retains itspreferred position in rickettsial infections. In atrial in scrub typhus the order of efficiency waschloramphenicol, aureomycin, terramycin andpara-aminobenzoic acid (Bailey, Ley, Diercks,Lewthwaite and Smadel, 1951).Summary

I. A historical review enables the identificationof certain principles which appear to have in-fluenced the development of antibiotics and theirselection for clinical application.

2. By projecting these principles, in a generalway, into the future, it appears possible to indicatelines of development.

3. 'T'he peculiar pathogenesis of certain diseasescaused by parasites susceptible to chemotherapylimits the application of antibiotics and thus in-fluences their development.

4. The discovery of new antibiotics effectiveagainst viruses, the gram-negative rods andprotozoa may be foreseen.

5. The replacement of inefficient antibiotics bymore efficient members is inevitable. The re-quirements include bactericidal rather than bac-teriostatic action, insusceptibility to the effect ofnumbers of organisms and a minimal rate ofemergence of resistant strains.

6. The most recent antibiotics are discussedand some facets of their development and applica-tion explored.

BIBLIOGRAPHYALEXANDER, H. E. (1947), Adv. Pediat., 2, 121.ALEXANDER, H. E., LEIDY, G., and REDMAN, W. (1949),J. din. Invest., 28, 867.BARER, G. R. (I951), J. Gen. Microkiol., 5, r.BEYER, K. H., RUSSO, H. F., PATCH, E. A., TILLSON, E. K.,

and SHAVER, G. (1947), J. Pharmacol., 91, 272.BRAKLEY, E. (1950),.. Pediat., 37, 42.BROWNLEE, G. (195I), Ann. Rev. Microbiol., 5, I97.BROWNLEE, G., and BUSHBY, S. R. M. (1948), Lancet, i, 127.BROWNLEE, G., BUSHBY, S. R. M., and SHORT, E. I.,

(1949), Ann. N.Y. Acad. Sci., 51, 952.BROWNLEE, G., BUSHBY, S. R. M., and SHORT, E. I. (1949).

Brit. J. Pharmacol., March 1952 (in press).BRYER, M. S., SCHOENBACH, E. B., and BLISS, E. A.

(I949),-Ann. N.Y. Acad. Sci., 5x, 935.CHAIN, E., FLOREY, H. W., GARDNER, A. D., HEATLEY,

N. G., JENNINGS, M. A., ORR-EWINGS, J., andSANDERS, A. G. (1940), Lancet, ii, 226.

CARR, D. T., PFUETZE, K. H., BROWN, H. A., DOUGLAS,B. E., and KARLSON, A. G. (I95I), Amer. Rev. Tuberc,,63, 427.

CHANDLER, C. A., and HODES, H. L. (1950), Pediatrics, 5, 267.COLEBROOK, L., and KENNY, M. (1936), Lancet, ii, 1319.DEMEREC, M. (1945), Proc. Nat. Acad. Sci. Wash., 31, 16.DOMAGK, G. (I935), Deutsch. Med. Wschr., 6x, 829.

ELIAS, W., PRICE, A. H., and MERRION, H. J. (i95I),Antibiotics and Chemotherapy, I, 491.

FINLAY, A. C., HOBBY, G. L., HOCHSTEIN, F., LEES,T. M., LENERT1 T. F., MEANS, J. A., P'AN, S. Y., REGNA,P. P., ROUTIEN, J. B., SOBIN, B. A., TATE, K. B., andKANE, J. H. (i951), Amer. Rev. Tuberc., 63, I.

FLEMING, A. (I929), Brit. J. exp. Path., xo, 226.FRANCIS, J., SPINKS, A., and STEWART, G. T. (I950), Brit.

J. Pharmacol., 5, 549.HALL, E. H., KAVANAGH, F., and ASHESHOV, I. N. (I95x),

Antibiotics and Chemotherapy, I, 369.HAYES, E. R., and YOW, E. (1950), Amer. J. Med. Sci., 22o, 633HENRY, R. J., and HOBBY, G. L. (I949), ' Streptomycin, Nature

and Practical Applications,' S. J. Waksman, Williams Wilkins,Baltimore.

KIKUTH, W. (1932), Deutsch. med. Wschr., 58, 530.LINZ, R. (I948), Compt. rend. Soc. Biol., x42, xo66.MILLER, C. P., and BOHNHOFF, M. (I946), J.A.M.A., 130, 485RAKE, G., and RICHARDSON, A. P. (1946), Ann. N.Y. Acad.

Sci., 48, 143.RAMMELKAMP, C. H., and BRADLEY, S. E. (1943), Proc.

Soc. exp. Biol., N. Y., 53, 30.SALLE, A. J. (1950), ' Bacteriological Proceedings,' p. 72, M. 8.SCHOENBACH, E. B., BRYER, M. S., and LONG, P. H. (1949),

Ann. N.Y. Acad. Sci., 5x, 987.(Continued on page 157)

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from

March 1952 SNELLING AND CONES: Malignant Tumours of the Testis 157

ANNOTATION

The Bone BankIn many centres in this country a ' bone bank'

is becoming a standard part of the facilities of anorthopaedic department. The development of thisservice has ,taken place mainly in the UnitedStates, particularly by P. D. Wilson (I95I).Bone grafting is an everyday procedure where

sound and rapid bone healing is required. Bonetaken from the patient-an autogenous graft-hasno blood supply, and all the cells must die, exceptperhaps those on the very surface. Its use is togive mechanical stability while new bone is formedin and around the graft, which itself becomesincorporated and replaced. This is the sameprocess which takes place when the small frag-ments of a comminuted fracture become embeddedin a mass of callus as the fracture heals.

Experimental work would suggest that a homo-genous graft (from another patient) differs froman autogenous graft only in the invariable deathof the surface, as well as of the deep cells. Thisseems to be unimportant, and the final results areequally satisfactory whichever type of graft is

used. Thin slices of cancellous bone probablysurvive complete as living entities and such graftslaid, for instance, on a skull defect becomeclinically sound within three weeks.The bone bank consists of a deep-freeze re-

frigerator kept at a constant temperature of-20° C.The bone is usually from the rib of a young patientundergoing thoracotomy for a non-infective con-dition, a Wassermann reaction being taken beforethe bone is used. The rib taken at operation isplaced in a sterile screw-topped jar containingnormal saline in 'which is dissolved penicillin(o0,000 units) and streptomycin (Io,ooo micro-grarms). Most authorities credit this procedureas being satisfactory in preserving homogenousgrafts for at least four months.

Recently, Kreuz, et al. (195I), have used bonewhich has been freeze-dried by the techniquesuggested by Flosdorf (I949) and then stored atroom temperature. This technique is particularlysuitable for isolated units where a suitable' re-frigerator is riot available. It has been suggestedthat this technique should be used under war-timeconditions.

J.S.R.G.BIBLIOGRAPHY

FLOSDORF, E. W. (1949), 'Freeze Drying by Sublimation,'Reinhold Publishing Corp., p. 138.

KREUZ, et al. (1951), J. Bone Jt. Surg., 33, 863.WILSON, P. D. (I951), Ibid., 33, 301.

NOTENEW PRODUCT-' DIAMETHINE'

Burroughs Wellcome & Co. announce the introduction of' Diamethine' brand injection of dimethyl-tubocurarine bromide, a muscle relaxant with an action similar to, but of slightly shorter duration than,d-tubocurarine chloride, and causing less histamine release. 'Diamethine' is the bromide salt of thedimethyl ether of d-tubocurarine, and is miscible in all proportions with intravenous barbiturates.Intended as a companion product to ' Tubarine' brand injection of d-tubocurarine chloride, it is issuedin a strength (4 mgm. per c.c.) approximately equipotent in muscle-relaxing effect with 'Tubarine.''Diamethine' is available as ampoules containing 6 mgm. in 1.5 c.c. (box of 6, 2Is. 6d.; 25, 83s. 6d.,subject) and rubber-capped bottles of 20 mgm. in 5 c.c. (Ios. 9d. each, subject).

(Continuedfrom page 144)SILVER, H. K., and KEMPE, G. H. (x947), J. Immunol., 57, 263.SPAULDING, E. H., BONDI, A., and EARLY, E. (1947), J. Lab.

din. Med., 32, 807.SWIFT, P. N., and BUSHBY, S. R. M. (1951), Lancet, ii, 183.WAKSMAN, S. H., and LECHEVALIER, H. A. (1949), Science.

109, 305.Aureomycin, Chloramphenicol and Terramycin:ARMSTRONG, T. G., WILMOT, A. J., ELSDON-DEW, R.

(1950), Lancet, ii, Io.BAILEY, C. A., LEY, H. L., DEERCKS, F. H., LEWTHWAITE,

R., and SMADEL, J. E. (I95I), Antibiotics and Chemotherapy,I, 16.

HERREL, W. E., HEILMAN, F. R., and WELLMAN, E. E.,Ann. N. Y. Acad. Sci., 53, 448.

HOBBY, G. L., LENERT, T. F., DONIKIAN, M., and PIKULA,D. (I95I), Amer. Rev. Tuberc., 63, 434.

LEVADITI, C., and VAISMAN, A. (195'), Chemotherapy andAntibiotics, I, 425.

LEY, H. L., SAYER, W. J., HOBSON, A. C. S., VANREENEN,R. M., TIPTON, V. J., FRICK, L. P., BALLARD, E. L.,TRAUB, R. (i95I), Antibiotics and Chemotherapi, r, 28I.

LOUGHLIN, E. H., and JOSEPH, A. A. (I95I), Ibid., 1, 76.SIGUIER, F., and CHOUBRAC, M. (ig95), Bull. Soc. mdd, H6p.de Paris, 67, 217.

copyright. on M

ay 21, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.28.317.139 on 1 M

arch 1952. Dow

nloaded from