J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

J. clin. Path. (1964), 17, 504

The evaluation and modification of a technique for comparing the efficiency of antiseptics against subcutaneously deposited bacteria in mice

LEONARD JEFFRIES AND S. A. PRICE Walton Oaks Experimental Station, Dorking Road, Tadworth, Surrey

SYNOPSIS An attempt has been made to study the action of antiseptics on artificial streptococcal 'wounds' in mice by a published technique (Martin, 1959). The results were not in agreement with those of Martin and the possible reasons for this are discussed. Suggestions are made for modi- fying the original method to increase its scope and precision. The production and treatment of Ps. pyocyanea and K!. aerogenes 'wounds' in mice is described and discussed.

Martin (1959) described a method for studying the EXPERIMENTAL MATERIALS action of antiseptics on subcutaneously deposited streptococci in mice. In the basic technique mice ANAESTHETICS Two were tested. were anaesthetized with Avertin and a dose of Tribromethanol solution Avertin (Bayer Products Ltd.,

Kingston-on-Thames) was diluted 1 in 40 in 10% ethylcopyright. Strep. pyogenes, sufficient to cause death in untreated alcohol to give a 2-5 % solution immediately before use, mice, was injected subcutaneously; the needle was and tested with congo red to ensure that it had not left in position and the antiseptic was injected decomposed. through it into the same site 20 minutes later. Pentobarbitone sodium, B.P. solution Nembutal, Efficiency ofthe antiseptic wasjudged by comparison veterinary (Abbott Laboratories Ltd. Queensborough, of mortality rates and mean survival times of treated Kent), was diluted 1 in 10 in isotonic saline, to give a mice which died with that of untreated controls. 0-6% solution, which was freshly prepared before the In some experiments attempts were made to simu- start of each experiment. http://jcp.bmj.com/ late natural wound conditions by suspending both MICE A closed colony of randomly-mated albino mice the streptococcus and antiseptic separately in horse has been maintained for four years from stock supplied blood. When antiseptics were dissolved in horse by Mr. F. Sullivan, Pharmacology Department, Guy's blood at the concentrations recommended by their Hospital Medical School. Mice within the weight range respective manufacturers for the treatment ofwounds of 17 to 20 g. were used. chlorhexidine was stated by Martin to be markedly effective when compared with four quaternary ANTISEPTICS Solutions or suspensions were prepared in ammonium compounds, the effects of which were 90% defibrinated horse blood or water by adding 1 on September 28, 2021 by guest. Protected to be Of 14 'conventional volume of aqueous antiseptic solution at 10 times the judged slight. antiseptics', required strength to 9 volumes of blood or water. Of presumably dissolved or suspended in water, eight these, 0-2 ml. was administered subcutaneously in all were without effect, two were slightly effective, experiments. The following antiseptics were tested: three were doubtful, and only one (acriflavine 0 1 %) Chlorhexidine, phenoctide, domiphen bromide, benz- was clearly active. alkonium chloride, dequalinium chloride, and compound When we attempted to screen compounds 1162 (synthesized by, and under experimental investi- synthesized in these laboratories by this method it gation in, these laboratories). became clear that there were complicating factors leading to death of mice such as anaesthetic and INFECTING ORGANISMS Cultures were diluted in either toxicity. In our investigations we used a nutrient broth or 90% defibrinated horse blood, according antiseptic to the experiment, and injected subcutaneously in a different strain of mouse and Streptococcus haemo- volume of 0-05 ml. lyticus of a different Lancefield group from those Streptococcus haemolyticus, group C(Pion strain, C.N.4. used by Martin. Wellcome Collection of Bacteria). This strain was main- Received for publication 10 January 1964. tained by regular passage through mice at weekly intervals. 504 J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

Evaluation and modification technique for comparing antiseptics against subcutaneous bacteria in mice 505 Heart blood from a moribund infected mouse was with Strep. haemolyticus and treated with 0-1 % inoculated into brain-heart infusion broth (Difco) chlorhexidine, 0-1% compound 1162, or 0-5% containing 10% defibrinated horse blood and incubated phenoctide the mortality rates were high in all overnight. On the following day 1 ml. was added to 10 ml. groups whether or not 90% horse blood was used brain-heart infusion broth containing 10% normal horse for the and serum, and incubated for six hours. For use the culture suspending antiseptics bacteria. All mice was diluted 1 in 1,000; this produced a fatal septicaemia in the saline-treated infection control groups died in mice in 2 to 2 5 days. and so also did those treated with 0-5 % phenoctide. Klebsiella aerogenes (N.C.T.C. 5055) This organism The phenoctide-treated mice in fact died earlier than was deposited in the National Collection of Type the saline-treated controls, but whereas Strep. Cultures in 1937 as Klebsiella pneumoniae. In accordance haemolyticus was recovered from the heart blood of with the classification proposed by Cowan, Steel, Shaw, all the control mice it could not be recovered from and Duguid (1960) it is now regarded as a strain of most of the mice that had been treated with Klebsiella aerogenes (Steel, 1963). An overnight brain- phenoctide. It was apparent therefore that the heart infusion broth culture diluted 1 in 1,000 regularly been caused a septicaemia in mice and was fatal in an average deaths may have due to the toxicity of this time of 3-5 days. The virulence of this strain remained antiseptic which nevertheless killed the streptococcus constant over seven months without 'mouse passage'. at the injection site. Subsequent experiments with Pseudomonaspyocyanea Two laboratory stock strains, unanaesthetized mice have confirmed that 0 5 % one of which, W.O. 83, was originally isolated from a phenoctide is indeed toxic and that anaesthetized human wound infection, and W.O. 126 isolated from mice are killed by even lower concentrations. a case of bovine mastitis, were chosen. Scrutiny of Martin's results reveals that he too A six-hour brain-heart infusion broth culture of either found the survival times of infected mice treated strain diluted 1 in 10 in horse blood produced a severe with 0-5 % phenoctide to be shorter than those of abscess at the injection site within six days in the majority of mice and a few mice died from septicaemia before the infected controls. He assumed that deaths were due sixth day. The injection of fewer organisms resulted in to infection and concluded that the activity of lesions in only a small proportion of mice. Two other phenoctide, at the recommended level of 0 5 %, was strains of Ps. pyocyanea were found to be unsuitable as very slight. We consider that the antiseptic is toxic their virulence for mice, by the subcutaneous route, was at that level and that it is necessary to test at lower copyright. too low. levels in order to demonstrate its antibacterial activity uncomplicated by toxicity. THE TECHNIQUE OF SUBCUTANEOUS We found that Avertin solution, diluted as INJECTION AND TREATMENT WITH ANTISEPTICS recommended by Martin and injected intra- During the periods of induction of anaesthesia and peritoneally in a dose of 0-2 ml., was toxic to recovery mice were placed on several layers of cellulose approximately 30% of our mice within the weight wadding in Harwell-pattern cages; shallow trays, range 17-22 g. Ten per cent of mice died within six similarly lined, were used in the actual experiments. hours of injection and a 20 % http://jcp.bmj.com/ Groups of mice were anaesthetized by the intraperitoneal further regained injection of either Avertin of Nembutal. When all mice consciousness but subsequently died during the in a group were completely relaxed each animal was following five days. placed on its right side and rested against the inner edge From Martin's original description and discussion of the tray. this ingenious technique appeared to represent a Each mouse was injected subcutaneously in the region potential advance on previously described methods. of the left shoulder with 0-05 ml. of bacterial suspension We found, however, that toxicity, both of anaesthetic through a gauge 25 hypodermic needle attached to a and antiseptic, may be serious complicating factors on September 28, 2021 by guest. Protected tuberculin syringe. With care to avoid altering the that may readily vitiate results. position of its point in the subcutaneous tissues the needle was detached from the syringe and the hub was rested on the body of the mouse. Twenty minutes later MODIFICATIONS TO THE BASIC TECHNIQUE 0-2 ml. of antiseptic solution was injected through the same needle, which was then removed. In the earlier CHOICE OF ANAESTHETIC In his description of experiments one group of control mice was used; these Avertin anaesthesia Martin stated: 'This anaesthetic were infected and subsequently treated with 90% horse was found superior to pentobarbitone for this blood or saline. Later experiments included an additional purpose, because deaths from the anaesthetic did group of mice which were also injected with 90% horse not occur'. blood as a control of anaesthetic toxicity. Some strains of mice are known to be more EXPERIMENTS AND RESULTS sensitive to Avertin than to Nembutal (pento- barbitone) and vice versa and, for this reason, EXPERIENCES WITH THE TECHNIQUE AS DESCRIBED BY different workers favour one or other anaesthetic MARTIN In Avertin-anaesthetized mice infected (Brown, 1963). It is possible that some of the deaths J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

506 Leonard Jeffries and S. A. Price in the experiments reported by Martin may also have all be safely administered to Nembutal-anaesthetized been due to delayed toxicity of Avertin. mice at the following concentrations (the con- Nembutal solution diluted 1 in 10 in saline was centrations recommended for wound treatment are given intraperitoneally in doses from 0 15 ml. to shown in brackets):- Chlorhexidine diacetate, 0-3 ml. to groups of mice of both sexes in the weight 0 5% (0-05 %), phenoctide, 0-25 % (0-5 %), domiphen range 17-20 g. No sexual differences in reaction to bromide, 0'5% (0-5 %), benzalkonium chloride, the anaesthetic were noted. With doses of less than 0-1 % (01 %), dequalinium chloride, 0 4 % (0 4 % in 025 ml. the depth of anaesthesia, as judged by 'tulle gras'), and compound 1162, 0 5%. degree of relaxation and reaction to a needle prick of the skin, was unsatisfactory. A dose of 0-3 ml. RECOVERY OF INFECTING ORGANISM FROM MICE All was toxic to 30 % of mice within six hours of dosing; infected mice dying during the course of experi- mice surviving this time were still alive at the ments were examined after death and heart-blood end of the nine-day observation period. cultures were made on blood agar. Further cultures, From these preliminary experiments a dose of as described below, were prepared from surviving 0-25 ml. Nembutal (approximately 75 mg./kg.) mice infected with Ps. pyocyanea. appeared to be satisfactory and this dose was used to anaesthetize 1,000 mice used in subsequent APPLICATION OF THE MODIFIED TECHNIQUE TO THE infection and toxicity experiments. In most experi- EVALUATION OF ANTISEPTICS AGAINST THREE ments a group of anaesthetic control mice was used. DIFFERENT INFECTING ORGANISMS These were anaesthetized and 30 minutes later were injected with 0-2 ml. of 90% horse blood in saline. STREPTOCOCCUS HAEMOLYTICUS In the experiments Altogether 90 mice were used for this purpose; recorded in Table I antiseptics and streptococcal 3 4% of these died within six hours of anaesthesia suspension were prepared in 90% horse blood. The and the remainder survived the duration of the antiseptics were administered 20 minutes after experiment. infection and mice were observed for nine days. It is essential to test anaesthetic toxicity on the Septicaemia was confirmed as the cause of death strain of mouse to be used in this technique; the by recovery of the streptococcus from the heartcopyright. mice should be observed for nine days. It is also blood of all mice that died. The number of mice in recommended that an anaesthetic control group of each group varied from one experiment to another mice should be included in each experiment involving due to supply. Variation in the number of mice in anaesthesia. the groups within an experiment was due to anaes- thetic deaths and technical failures, such as blockage ANTISEPTIC TOXICITY TESTS IN MICE ANAESTHETIZED or misplacement of the hypodermic needle. Mice in of seven to 10 Table 1, WITH NEMBUTAL groups I, experiment shows that chlorhexidinehttp://jcp.bmj.com/ were anaesthetized with Nembutal and 30 minutes and domiphen bromide were active at the recom- later they were injected subcutaneously with 0 2 ml. mended user concentration and dequalinium of antiseptic solution or suspension in 90% horse chloride was active at the concentration in which it is blood. As a result of these experiments it was used in a mixture (Wilkinson, 1959). Compound established that the antiseptics to be studied could 1162 was active at an arbitrarily selected concentra- BLE I THE EFFECT OF ANTISEPTICS USING STREP. HAEMOLYTICUS AS TEST ORGANISM on September 28, 2021 by guest. Protected Experiment Treatment Mortality Mean Survival Time (days) of Mice Dying Chlorhexidine, 0-1 % 1/7 (14%) 3 0 Chlorhexidine, 0 05 %1 0/9 Compound 1162, 01 % 1/8 (12%) 20 Benzalkonium chloride, 0-1 % 8/9 (89%) 3.4 Phenoctide, 0-25 %2 2/9 (22%) 3 0 Domiphen bromide, 0-5 %1 1/8 (12%) 3-0 Dequalinium chloride, 0-1 % 3/9 (33%) 3-3 Blood saline (infection control) 7/7 (100%) 2-6 2 Chlorhexidine, 0-01 % 8/8 (100%) 3-4 Dequalinium chloride, 0-01 % 3/9 (33%) 2-7 Compound 1162, 0-01 % 6/8 (75%) 5.5 Blood saline (infection control) 8/8 (100%) 2-6 'Manufacturer's recommended strength for wound treatment. 2Highest concentration tolerated by mice. Nine mice in each group at start of experiment. Antiseptics and organisms were suspended in blood. J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

Evaluation and modification technique for comparing antiseptics against subcutaneous bacteria in mice 507 TABLE II EFFECT OF ANTISEPTICS USING STREP. HAEMOLYTICUS AS TEST ORGANISM' Experiment Treatment Mortality Mean Survival Time (days) of Mice Dying 1 Domiphen bromide, 0 5 % 1/8 (12%) 1*0 Domiphen bromide, 0-1 % 1/8 (12%) 40 Phenoctide, 0-1 % 1/8 (12%) 30 Blood saline (infection control) 8/8 (100%) 2-5 2 Chlorhexidine, 0-1 % 1/10 (10%) 10 Chlorhexidine, 0-01 % 10/10 (100%) 2-8 Phenoctide, 0-01 % 10/10 (100%) 2-8 Domiphen bromide, 0-01 10/10 (100%) 2-3 3 Benzalkonium chloride, 0-1 4/9 (44%) 3-25 Compound 1162, 0-1 % 0/9 Compound 1162, 0-01 % 9/9 (100%) 3-66 Blood saline (infection control) 10/10 (100%) 2-2 'Antiseptics were dissolved in blood, and streptococcus was suspended in broth.

TABLE III INFLUENCE OF BLOOD AND BROTH AS SUSPENDING AGENTS FOR THE STREPTOCOCCUS ON THE ACTIVITY OF CERTAIN ANTISEPTICS DISSOLVED IN BLOOD Treatment Medium for Suspension of Mortality Mean Survival Time (days) of Streptococcus Mice Dying Phenoctide, 0-1 % Blood 9/10 (90%) 28 Phenoctide, 0-1% Broth 3/10 (30%) 23 Domiphen bromide, 0-1 % Blood 7/10 (70%) 2 3 Domiphen bromide, 0 % Broth 1/10 (10%) 20 Blood saline (infection control) Blood 10/10 (100%) 22 copyright. tion and benzalkonium chloride showed only bromide, benzalkonium chloride, and compound slight activity. Phenoctide was active at the maximum 1162 solutions at 0 1 % in blood were tested against tolerated concentration. the streptococcus suspended in broth all five were In Table I, experiment 2, three of the compounds active. When the same concentrations of antiseptics shown to be active in experiment 1 were diluted in blood were tested against the streptococcus further (five to ten fold) and, judging from mortality suspended in blood the activities of phenoctide,

rates and mean survival times, compound 1162 and domiphen bromide, and benzalkonium chloride http://jcp.bmj.com/ dequalinium chloride appeared to be more active were reduced. It would therefore appear that than chlorhexidine. streptococci suspended in blood are less accessible Whether or not it would be possible to distinguish to certain antiseptics although these antiseptics are between the activities of compound 1162 and also suspended in blood; it is not clear to us why dequalinium chloride by further tests against the this should be. streptococcus has not been investigated. Martin's conclusion that, at the recommended It was found that, when the streptococcus was user levels, chlorhexidine was more effective than on September 28, 2021 by guest. Protected suspended in broth instead of 90% horse blood, the phenoctide and domiphen bromide may be errone- needle became blocked less frequently. Martin ous, as he failed to exclude anaesthetic and anti- found that the mean survival time of infected, septic toxicity, nor did he state that he had untreated mice was the same whether blood or established infection as the cause of deaths. broth was used for suspending the streptococcus; By the modified technique, in which all deaths we have confirmed this. The results in Tables I and were confirmed as streptococcal, we found domiphen II, together with those of other experiments for bromide to be active at the recommended user level. which there is not space to detail, suggested that Phenoctide was active at a lower level than that certain antiseptics were adversely affected by the use tested by Martin. of blood instead of broth for suspending the streptococcus. Table lII, in which two quaternary PSEUDOMONAS PYOCYANEA Ps. pyocyanea appears to ammonium compounds were compared at one be of low virulence for mice by the subcutaneous fifth the recommended user levels in one experiment, route. We found that the subcutaneous injection of illustrates this effect. large numbers of cells of the more virulent strains When chlorhexidine, phenoctide, domiphen caused septicaemia fatal within six days in only a J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

508 Leonard Jeffries and S. A. Price small proportion of mice; the remainder developed It is of interest to note that, although all infected, local abscesses. Two such strains were selected for untreated control mice showed local abscesses due to this investigation. Ps. pyocyanea when theywere killed, only oneout ofa Groups ofmice were anaesthetized with Nembutal total of 28 such mice (Table IV) died from septic- and injected subcutaneously with 005 ml. of a aemia. In contrast, a high proportion of mice six-hour Ps. pyocyanea culture diluted 1 in 10 in treated with ineffective quaternary ammonium either broth or 90% horse blood. The antiseptic compounds died from septicaemia. Phenoctide and solution in horse blood was injected into the same domiphen bromide, although clearly ineffective at site 20 minutes later. Cultures were prepared from high concentrations, were tested at a single lower the heart blood of any mice that died during the level and the proportion of mice dying from following six days; survivors were then killed and septicaemia was not reduced; still lower doses were the tissues at the site of injection were dissected. not tried. These observations are presumably related Lesions were recorded and any contents were to the mechanism of drug toxicity on the infected examined by Gram's stain and inoculated on to mouse, and certain observations of Dutton (1955) agar plates, which were incubated for three days. may be relevant. Most antiseptics, in the concentrations tested, Dutton (1955), in a study of the effect of the route produced slight tissue necrosis. Infected lesions, of injection on lethal infection of mice, found that however, were usually easily distinguished and eight species of bacteria were less lethal when confirmed by demonstration of pus cells and Gram- injected intravenously than by either the intra- negative bacilli microscopically and the growth of peritoneal or subcutaneous routes. He suggested Ps. pyocyanea in culture. In untreated mice abscesses that the more rapidly certain bacteria reach the started to resolve after the sixth day. blood stream (via macrophages and lymphatics) Although results were similar whether horse blood from a primary site the greater is the possibility of or broth was used for suspending the Ps. pyocyanea, the host's survival. Thus localization of the infecting blood was used in most experiments. The results of organism at the primary site is not necessarily the

three such experiments, using Ps. pyocyanea (strain most effective defence reaction. It may be thatcopyright. W.O. 83), are given in Table IV. certain antiseptics, such as phenoctide, interfere with It will be seen from Table IV that benzalkonium the mechanism for the removal of small numbers of chloride, domiphen bromide, and dequalinium Ps. pyocyanea from the subcutaneous site to the chloride, in the concentrations recommended for blood stream where they could be destroyed, and wound treatment and shown to be non-toxic to thereby permit the development of a focus of mice, and phenoctide, at the highest level tolerated infection of severe degree that ultimately over- by mice, did not suppress Ps. pyocyanea infection. whelms the host and produces a fatal septicaemia.

Chlorhexidine, at the user level of 0.05%, showed It may be possible to develop a method, based onhttp://jcp.bmj.com/ considerable protection and at 0-1 % the effect was this phenomenon, for studying the toxicity of marked. antiseptics to wound tissue. It is well known that the TABLE IV EFFECT OF ANTISEPTICS DISSOLVED IN BLOOD AGAINST PS. PYOCYANEA (sTRAIN w.o. 83) INFECTION

Experiment Treatment No. ofMice in Mortality from Ps. No. of Survivors with Y. of Mice that Suffered on September 28, 2021 by guest. Protected Group pyocyanea Septicaemia Lesions due to Ps. from Infection pyocyanea No. of Mice % of Total

Chlorhexidine, 01 % 10 0 10 Chlorhexidine, 0-05 %' 9 0 5 55 Blood saline (infection control) 11 1 9 10 100 Chlorhexidine, 01 % 9 0 0 0 Phenoctide, 0-25 % 10 10 100 100 Domiphen bromide, 05%' 9 4 44 5 100 Benzalkonium chloride, 0-1 %'. 10 7 70 2 90 Blood saline (infection control) 9 0 9 100 Chlorhexidine, 0-1 % 8 0 12 Compound 1162, 0-1 % 7 14 6 100 Dequalinium chloride, 0 4 %' 9 5 55 4 100 Phenoctide, 0-1 % 9 9 100 100 Domiphen bromide, 0-1 % 8 5 62 3 100 Blood saline (infection control) 8 1 12 7 100 'Manufacturer's recommended strength for wound treatment. J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

Evaluation and modification technique for comparing antiseptics against subcutaneous bacteria in mice 509 TABLE V EFFECI OF ANTISEPTICS AGAINST KL. AEROGENES INFECTION Experiment Treatment No. of Mice in Group Mortality from KI. aerogenes Mean Survival Time (days) Septicaemia of Mice Dying No. of Mice Y. of Total 1 Chlorhexidine, 0-1 % 10 10 50 Chlorhexidine, 0 0S %1 9 2 22 4.5 Domiphen bromide, 0-5 %/ 10 4 40 50 Blood saline (infection control) 11 11 100 38 2 Chlorhexidine, 0-1 % 11 0 Compound 1162, 01 11 11 100 363 Domiphen bromide, 0-5 %5, 9 5 55 40 Phenoctide, 0-25 % 11 11 100 254 Benzalkonium chloride, 011 10 10 100 3.9 Blood saline (infection control) 10 10 100 3.3 'Manufacturer's recommended strength for wound treatment. All mice were observed for nine days.

results of studies in vitro on leucocyte damage by principles of biological assay. The evaluation of an antiseptics may be considerably influenced by the antiseptic must depend, therefore, on the application techniques used and their relevance and repro- of carefully selected tests and a critical appraisal of ducibility have been questioned (Fleming, 1940; the results obtained rather than on a comparison of Garrod, 1955). one compound against another by any single With infection produced by another strain of laboratory procedure; on the basis of such evaluation Ps. pyocyanea (W.O. 126) seven out of10 mice treated the compounds most meriting clinical trial must be with 0-1 % chlorhexidine in blood and all 10 blood- chosen. saline treated mice showed abscesses on the sixth Although there is an extensive literature on the day. testing in vitro of wound antiseptics relatively few copyright. methods have been described for testing in animals. KLEBSIELLA AEROGENES Nembutal-anaesthetized Attempts have been made to assess antiseptic mice were injected subcutaneously with 0-05 ml. of activity in mice by intraperitoneal injection immedia- an overnight broth culture of Ki. aerogenes diluted ately before, or at various intervals after, infection 1 in 1,000 in 90 % horse blood; antiseptics, suspended with lethal inocula of bacteria by the same route in blood, were injected 20 minutes later. The results (Babbs, Collier, Austin, Potter, and Taylor, 1956; of two separate experiments are shown in Table V. Cox and D'Arcy, 1963; Davies, Francis, Martin, This organism was tested because it is a highly Rose, and Swain, 1954). Such procedures might be http://jcp.bmj.com/ mouse-pathogenic encapsulated strain and represent- criticized on the grounds that peristalsis may rapidly ative of a group of coliforms that are potential disperse intraperitoneally injected organisms pathogens of man. At the recommended user level, creating multiple foci of infection that may be against this Klebsiella strain, domiphen bromide inaccessible to the antiseptic. showed a moderate degree of activity and chlor- Methods involving the establishment of experi- hexidine showed a more marked effect. Phenoctide, mentalwounds have been discussed by Martin (1959).

at the maximum level that could be tested, and Such methods, as he points out, have failed to on September 28, 2021 by guest. Protected benzalkonium chloride, at user level, were without control all the variables influencing the result and effect. the indwelling needle technique should go far towards meeting this difficulty. The technique was DISCUSSION able to differentiate between compounds of marked inhibitory action in vitro although no great precision The difficulties inherent in the laboratory testing of could be claimed for it. Our results have confirmed antiseptics are well known and have been extensively that the test, with certain modifications, is of value reviewed by, for example, Garrod (1955) and Sykes for this purpose. (1958). The comparison, one with another, of The choice of concentration of antiseptic to be substances of widely different chemical structure tested presents some difficulty. In general, we having diverse modes of antibacterial action and a would recommend that the highest non-toxic wide variety of toxic effects on the host inevitably concentration should be used but it is imperative poses complex problems. Indeed, any attempt to that the non-toxic level should be established with compare, by tests on living systems, qualitatively some care-itself a difficult procedure that may be different substances is at variance with the established influenced by the choice of anaesthetic. If, under J Clin Pathol: first published as 10.1136/jcp.17.5.504 on 1 September 1964. Downloaded from

510 Leonard Jeifries and S. A. Price these conditions, antiseptic A proves 'better' than were equally sensitive to chlorhexidine in vitro, antiseptic B then could higher doses ofB be tolerated differ markedly in their susceptibility to the anti- and the rating reversed? Alternatively, if A is bacterial action of chlorhexidine when tested by the 'effective' at a markedly lower level than B is it indwelling needle technique. It would be interesting necessarily a 'better' antiseptic? Calman and to know what proportion of Psuedomonas pyocyanea Murray (1956), in a study of antiseptics for use in strains isolated from wound infections would prove obstetrics, pointed out the importance of the virulent for mice and show sensitivity to chlor- economic aspect in their practical evaluation. hexidine by this method. Furthermore, compound Effectiveness at low concentrations may tempt the 1162, which proved highly effective against the manufacturers of an expensive compound to streptococcus and superior to chlorhexidine when recommend the lowest concentration consistent tested at the very low level of 0-01 %, was ineffective with efficiency, but the principal advantage of better when Pseudomonas pyocyanea was used as the test antiseptics should be that they can safely be used at organism. more effective levels. When applied critically with due regard for the In the technique as originally developed by spurious effects that may be unwittingly introduced Martin Streptococcus pyogenes was the only through anaesthetic or antiseptic toxicity, and organism used for infection of the mice. We regard considered in conjunction with results of sensitivity the extension of the technique to include other test tests in vitro and contact time estimations, we are of organisms as a valuable modification. The number of the opinion that this technique can provide valuable bacterial species, however, that will produce either information on the relative merits of antiseptics and local or systemic infection in mice is small. In view of on their suitability for clinical trials. the importance of Staph. aureus in hospital infections at the present time, this would be a most realistic We thank Mr. W. A. Freeman, May & Baker Ltd., for test organism. Unfortunately, administered by the the strain of K!. aerogenes and Mr. C. D. Wilson, Central subcutaneous route to mice, it is well tolerated Veterinary laboratories, Weybridge, for one strain of Ps. pyocyanea. For advice on the anaesthesia of mice we (Smith, Wilson, Hazard, Hummer, and Dewey, 1960) are grateful to Dr. Annie Brown, Laboratory Animalscopyright. and we were unable, with any of six strains of Centre, Carshalton, and for information concerning the Staph. aureus, to produce systemic infections identity of the Kl. aerogenes strain we thank Dr. K. J. after subcutaneous injection of broth cultures. Steel, National Collection of Type Cultures. We are also We have, however, been able to produce local indebted to the manufacturers of some of the antiseptics abscesses by subcutaneous injection of suitable for providing us with samples of their products for strains, the severity and frequency of which could be investigation. enhanced by gastric mucin. Such abscesses were REFERENCES readily prevented by antiseptics. Further work is in http://jcp.bmj.com/ progress to determine whether the influence of Babbs, M., Collier, H. 0. J., Austin, W. C., Potter, M. D., and antiseptics on staphylococcal abscess formation Taylor, E. P. (1956). J. Pharm. Pharmacol., 8, 110. as a Brown, Annie (1963). Personal communication. could be used the basis of method for antiseptic Calman, R. M., and Murray, J. (1956). Brit. med. J., 2, 200. differentiation. Recent reports suggest that Ps. Cowan, S. T., Steel, K. J., Shaw, C., and Duguid, J. P. (1960). J. gen. pyocyanea infections are increasing in frequency Microbiol., 23, 601. Cox, W. A., and D'Arcy, P. F. (1963). J. Pharm. Pharmacol., 15, 129. and severity and, for these reasons, Ps. pyocyanea Davies, G. E., Francis, J., Martin, A. R., Rose, F. L., and Swain, G. (1954). Brit. J. Pharmacol., 9, 192. would appear to be an important organism to on September 28, 2021 by guest. Protected Dutton, A. A. C. (1955). Brit. J. exp. Path., 35, 128. include in antiseptic screening tests in mice. Fleming, A. (1940). Proc. roy. Soc. Med., 36, 487. Extension of the technique to include other Garrod, L. P. (1955). Ibid., 48, 21. has the not Martin, A. R. (1959). J. clin. Path., 12, 48. microorganisms emphasized importance Smith, I. M., Wilson, A. P., Hazard, E. C. Hummer, W. K., and only of species but also of strain differences in Dewey, M. E. (1960). J. infect. Dis., 107, 369. attempts to rank antiseptics for efficiency. Thus, Steel, K. J. (1963). Personal communication. Sykes, G. (1958). Disinfection and Sterilization. Spon, London. our two strains of Ps. pyocyanea, both of which Wilkinson, D. S. (1959). Practitioner, 182, 501.