Of Significant in Vitro Antagonism Between Penicillin, Cephalothin, and Rolitetracycline FRANZ D

ANTIMICROBIAL AGENTS AND CHEmoTHzRAPY, Nov. 1976, p. 802-808 Vol. 10, No. 5 Copyright © 1976 American Society for Microbiology Printed in U.S.A. Combination of Bacteriostatic and Bactericidal Drugs: Lack of Significant In Vitro Antagonism Between Penicillin, Cephalothin, and Rolitetracycline FRANZ D. DASCHNERI Children Hospital, Division ofAntimicrobial Therapy, University ofMunich, 8 Munich 2, Germany Received for publication 16 April 1976 Although it is generally believed that bactericidal and bacteriostatic drugs should not be combined in vivo, in vitro experiments using the checkerboard dilution technique revealed no antagonism between penicillin/cephalothin and rolitetracycline but rather additive or synergistic activity of either drug combination in 40 to 50% of20 Escherichia coli and 14 Staphylococcus aureus strains. Slight antagonism occurred only between 3 and 8 h after combining penicillin/ cephalothin and rolitetracycline in either bacteriostatic or bactericidal concentrations, but not after 24 h of incubation, nor was antagonism found with combinations of these drugs in bacteriostatic concentrations. Neither bacteriostatic nor bactericidal activity of penicillin/cephalothin and rolitetracycline was inhibited by pretreatment of one E. coli strain with bacteriostatic rolitetracycline or bacteriostatic penicillin/cephalothin concentrations. Penicillin and cephalothin could exert a bactericidal effect after 2-h exposure of theE. coli strain to bacteriostatic rolitetracycline concentrations. Combined action of subinhibitory penicillin and rolitetracycline concentrations resulted in more pronounced inhibition of growth than either drug alone. The higher activity of penicillin/ cephalothin in combination with rolitetracycline on some E. coli and S. aureus strains might be due to a better access of rolitetracycline into bacterial cells whose cell walls have been weakened by cell wall-active, bactericidal drugs. Thus, growth ofpenicillin-induced spheroplasts ofE. coli and stable staphylococcal L-forms was inhibited by much lower concentrations of rolitetracycline than were the corresponding parent cells with intact cell walls. According to Jawetz and Gunnison, the anti- streptomycin is not needed for effective treat- biotic and chemotherapeutic drugs can be di- ment of patients with S. viridans endocarditis vided into substances the effect of which is (2). Furthermore, it has long been known from mainly bactericidal (e.g., penicillin, cephalo- in vitro studies that decreasing or increasing sporins) and substances that generally have the concentration of primarily bactericidal and only bacteriostatic effects (e.g., tetracyclines) bacteriostatic drugs results in either bacterio- (11). They noted that whereas two bactericidal static or bactericidal activity (19). None ofthese drugs were often synergistic, a combination of studies, however, considered the original find- bactericidal and bacteriostatic antibiotics often ing of Jawetz that even within a given test resulted in antagonism in vitro. Several excep- system, the results of drug combinations may tions, however, to this chemotherapeutic rule depend on drug concentrations, time of interac- have been found. The bacteriostatic sulfona- tion, inoculum size, and other laboratory varia- mides may act synergistically with the bacteri- bles (10). I therefore investigated some of these cidal polymyxins against Proteus species and variables, especially subinhibitory, bacterio- Pseudomonas aeruginosa (20, 22). Antagonism static, and bactericidal drug concentrations and has been described between two penicillins, time of interaction, combining bactericidal ampicillin and carbenicillin, when tested (penicillin G, cephalothin) and bacteriostatic against Enterobacter cloacae (2). Although in- (rolitetracycline [RTCI) drugs in vitro. In fur- creased killing ofpenicillin-susceptible Strepto- ther experiments with cell wall-defective bacte- coccus viridans by penicillin plus streptomycin ria (L-forms, spheroplasts), I examined why the can be demonstrated in vitro, the addition of combination of bactericidal and bacteriostatic to instead of antago- I Present address: Wissenschaftlicher Rat und Professor, drugs may lead synergism Leiter der Klinkhygiene in der Universitets-Fraunklinik, nism (16). RTC was used because with this drug 7800 Freiburg, West Germany. very high serum levels can be obtained in pa- 802 VOL. 10, 1976 COMBINATION OF BACTERIOSTATIC/CIDAL DRUGS 803 tients after intravenous application, levels that the usual manner. The isobole defining additive exceed the concentrations found to be bacteri- effects of the two antibiotics was obtained by draw- cidal in this study (7, 15). ing a straight line between the minimal inhibitory concentrations (MICs) of each drug acting alone. MATERIAL AND METHODS Synergism would be evident if the line representing combined action of both drugs at various concentra- Test organisms. All Escherichia coli and Staphy- tions fell below the additive isobole (21). Since one- lococcus aureus strains used in this study were iso- tube variations in MICs are commonly observed lated from children with nosocomial diseases, with serial dilution tests, which have a marked mostly urinary tract infections and septicemia. One effect on the shape of the isobole, synergism was E. coli strain isolated from urine was used through- considered present only when there was, with the out the study for the growth and killing curves. The combination, at least a fourfold reduction in the four strains of stable staphylococcal L-forms (L-36, MICs of both antibiotics (e.g., a twofold reduction L-65, L-68, L-101) and the corresponding bacterial with one antibiotic and a fourfold with the other forms were obtained from B. M. Kagan, Cedars one, causing only a slight concave bowing of the Sinai Medical Center, Los Angeles, Calif., and sub- isobole was not accepted as evidence of synergistic cultured weekly as described previously (5). activity). A twofold reduction of the MICs of both or Antibiotics. Potassium penicillin G was supplied either was considered additive, and no lowering of by Bayer (Leverkusen), cephalothin by Eli Lilly & the end points was regarded as indifference. Co. (Bad Homburg), and RTC by Hoechst AG Growth curves. Growth curves (Fig. 4) of one E. (Frankfurt), each as standard powder. Laboratory coli strain in Mueller-Hinton broth or osmotically standard solutions were prepared by dissolving the stabilized brain heart infusion broth were obtained antibiotic in Mueller-Hinton broth; they were stored by measuring optical density, as described previ- at -22°C and thawed immediately before use. Fur- ously (5, 6). Subinhibitory penicillin (150 ,ug/ml) or ther dilution was done in Mueller-Hinton broth RTC concentrations (0.9 ,ug/ml) were added alone, (Merck AG, Darmstadt). in combination, or following pretreatment with one Media. Mueller-Hinton broth adjusted to pH 7.4 of these antibiotics after 3 h of incubation. or brain heart infusion broth (Difco) supplemented Killing curves. Serial colony counts were per- with 10% sucrose for osmotic stabilization was used formed with broth cultures of one E. coli strain to for the growth and killing curves. All synergy stud- determine the time-related bactericidal or bacterio- ies using the checkerboard dilution technique were static effects of penicillin or cephalothin alone or in done in freshly prepared Mueller-Hinton broth, ad- combination with RTC. Broth with one or both anti- justed to pH 7.4 and supplemented with phenol red as biotics was inoculated with bacteria from an over- indicator to facilitate reading the end poiints, as night culture and adjusted photometrically to con- described previously (4). Salt serum agar was pre- tain an average final inoculum of 0.4 x 106 orga- pared according to Kagan (12) using gamma globu- nisms/ml. At 0, 1, 2, 3, 4, 5, 6, and 24 h, 0.1 ml from lin-free horse serum (Grand Island Lab., N.J.) in- each tube was transferred undiluted or in 10-fold stead of human serum. dilutions in broth to petri dishes with Endo agar Synergy studies. The effect of RTC on the inhibi- (Difco). Colony-forming units per milliliter were tory activity of penicillin G or cephalothin was ob- counted after overnight incubation. Typical experi- served using a microtiter variation of the checker- ments are shown in Fig. 1 through 3. board dilution technique, as described previously Antibiotic susceptibility testing of L-forms. Anti- (4). Briefly, twofold dilutions of RTC in 0.025 ml biotic susceptibility testing was done by using the were made in one direction with a microtiter dilutor. agar dilution technique. From salt serum agar with Twofold dilutions of penicillin or cephalothin were equal growth of single L-form colonies, 0.5- by 0.5- added in the same volume manually in the perpen- cm agar blocks were cut out and inversely moved to dicular direction, using a disposable pipette deliver- one-half of the salt serum agar plates containing ing 0.025 ml/drop. Where each ofthe antibiotics was twofold dilutions of RTC (range, 0.45 to 1,000 ,ug/ used alone, 0.025 ml ofbroth was added to achieve a ml). The other halves of the plates were inoculated volume of 0.05 ml. The inoculum of 0.05 ml from an with 0.005 ml of a standardized inoculum (approxi- overnight culture, adjusted to contain approxi- mately 5 x 105 organisms) of the corresponding par- mately 106 organisms/rnl with a Coleman Junior ent forms, which resulted in growth of single colo- spectrophotometer, was added by delivering 1 drop nies. One plate was used to test all four L-forms and from a 0.05-ml disposable drop pipette. The final parent forms; one plate without antibiotics served as concentrations of RTC forE. coli (S. aureus ) ranged growth control. Plates were read after 24 h and from 0.09 to 12.5 ,ug/ml (0.02 to 3.12 ,tg/ml), those for again after 48 and 72 h. Antibiotic concentrations cephalothin ranged from 0.9 to 125 gg/ml (0.006 to that inhibited visible growth of L-forms and parent 0.9 ,ug/ml), and those for penicillin ranged from 117 forms of staphylococci were taken as the MICs. to 15,000 ,ug/ml (6 to 768 Mg/ml). Inhibitory end Incubation. All experiments were done at least in points were determined after overnight incubation triplicate under aerobic conditions at 37°C. at 37°C by noting the lowest concentration of antibiotics, alone and in combination, that allowed no RESULTS visible growth. This was greatly facilitated by change of the indicator from red to yellow. Isobolo- Synergy studies. Table 1 summarizes the ef- grams for each strain were constructed by plotting fects of penicillin and cephalothin in combina- the inhibitory end points on an arithmetic scale in tion with RTC on 20 E. coli and 14 S. aureus 804 DASCHNER ANTIMICROB. AGENTS CHEMOTHER. TABLE 1. Effect ofcombinations ofpenicillin (Pen)l therefore studied the effect of penicillin and cephalothin (Ceph) and RTC on 20 E. coli and 14 S. RTC on E. coli cells that had been pretreated aureus strains as determined by using the with either drug in bactericidal or bacterio- checkerboard dilution technique static concentrations. Neither bacteriostatic No. of strains nor bactericidal activity of penicillin and RTC Effect of anti- was inhibited by pretreatment with bacterio- biotic combina- E. coli S. aureus tion static RTC or penicillin concentrations (Fig. 3). Pen/RTC CephlRTC Pen/RTC Ceph/RTC In this type ofexperiment penicillin could exert Indifferent 10 10 8 10 a prompt bactericidal effect even after 2 h of Additive 10 7 6 3 exposure of the E. coli strain to bacteriostatic Synergistic 0 3 0 1 RTC concentrations (7.8 ug/ml). Killing curves almost identical to those shown in Fig. 1, 2, and 3 were obtained if cephalothin instead of peni- strains. Three E. coli and one S. aureus strain cillin was used in bactericidal (15.0 ,ug/ml) or were inhibited synergistically by the cephalo- bacteriostatic (5.0 Ag/ml) concentrations. Con- thin/RTC combination; penicillin/RTC or ceph- trol tubes without antibiotics were assayed in alothin/RTC combinations were found to be ad- all experiments to be sure that bacteria studied ditive on at least one-half of the strains tested. were always in the logarithmic growth phase. More important, however, was the lack of true All control curves were almost identical to the antagonism of combinations of any of the so- one shown in Fig. 1. called bactericidal and bacteriostatic antibiot- Growth curves ofE. coli and antibiotic sus- ics used in this study. The mean MIC ofRTC for ceptibility testing of staphylococcal L-forms. the E. coli (S. aureus) strains was 1.61 (0.69) The effect ofsubinhibitory RTC (0.9 ,ug/ml) and 4g1ml, and that ofpenicillin was 1,374 (161) ,uLg ml; the MIC of cephalothin was 15.6 (0.28) gg/ ml. Effect of bacteriostatic/bactericidal penicillin or RTC concentrations alone and in combination (killing curves). Figure 1 shows that significant growth over 24 h of the E. coli strain used in this study could be inhibited by certain penicillin (300 ,g/ml) or RTC (7.8 gg/ ml) concentrations (bacteriostatic concentrations). Increasing the concentrations resulted in immediate and marked bactericidal activity of both drugs. Only slight regrowth occurred beyond 8 h and up to 24 h. Figure 2 summarizes the findings if bacteriostatic/bactericidal penicillin/RTC concentrations or one drug in a bactericidal and the other in bacteriostatic concentration were combined at the beginning of the experiment. Control curves without antibiotics were similar to the one shown in Fig. 1. No antagonism could be found with penicillin/RTC combined in bacteriostatic concentrations (7.8 gg of RTC and 300 ,g of penicillin per ml). Slight antagonism occurred between 3 and 8 h after combining RTC or penicillin in either bacteriostatic or bactericidal concentrations. No antagonism, however, could be found after 24 h of incubation. It is well known that penicillin can destroy bacteria only when they are in the growing stage, so that drugs causing bacteriostasis might render penicillin ineffective if adminis- 1 2 3 4 5 6 7 8 24 tered simultaneously. In experimental infec- FIG. 1. Timed killing curves of one E. coli strain tions in mice, antagonism tended to occur espe- by bacteriostaticlbactericidal penicillin (3001600 pgl cially if therapy with chlortetracycline was ml) or RTC (7.8115.6 mg/ml) concentrations as com- started before therapy with penicillin (14). I pared to a control curve without antibiotic. VOL. 10, 1976 COMBINATION OF BACTERIOSTATIC/CIDAL DRUGS 805 CFU/mI spheroplasts by penicillin, which has been dem- 1O81 onstrated previously (6) as well as in this study by phase-contrast microscopy. This suggested that cell wall alterations induced by penicillin, especially in hypertonic medium, might permit greater access of RTC into the bacterial cell, which thus better reaches its site ofaction. This 106 assumption could be supported by my finding that growth of four different strains of stable staphylococcal L-forms (cell wall-deficient bacterial forms) was inhibited by significantly lower concentrations of RTC than was that of the corresponding parent forms with intact cell walls (Table 2). DISCUSSION Despite the theoretical problems posed by the *7.80pg/ml RTC+ 300pg/mI PEN use of antibiotic combinations, there are five 102- RTC+ PEN possible justifications for their use in clinical 07.80pg/mI 6OOpg/mI practice (2): (i) the concept of proven clinical _ 600 pg/mi PEN+15.60pg/mi RTC synergy; (ii) to broaden the spectrum; (iii) to 10~- treat mixed or potentially mixed infections; (iv) *150 g /ml RTC + 3001jg/mI PEN to prevent the eme'rgence of resistant organisms; and (v) the use of two potentially toxic drugs in lower dosage to possibly reduce the 1 2 3 4 5 6 7 8 24 FIG. 2. Timed killing curves ofE. coli by penicil- linIRTC combinations in either bacteriostatic or bac- * 7.80gg/mi RTC + aftor2h300pg/mi PEN tericidal concentrations. & 300 j/mi PEN + after2h780pqg/mt RTC CFU/mI * OOOpg/mi PEN + after2hlS60.pg/mi RTC penicillin (150 jig/ml) concentrations on the * 7.80ag/mi RTC + after2h 600 PEN growth of E. coli was studied by measuring 1o8I pg/mi optical density at hourly intervals. The curves obtained for growth~~~~~~~~~hin Mueller-Hinton broth are shown in Fig. 4, and those obtained for growth in osmotically stabilized brain heart infusion broth are shown in Fig. 5. Combined action of subinhibitory penicillin and RTC concentrations invariably resulted in more pronounced inhibition of growth than with either drug alone. Addition of penicillin or RTC 3 h after pretreatment with the other drug, each in a subinhibitory concentration, led to growth retardation; the growth curves were parallel, approaching the rate of growth inhibition ofE. coli by penicillin and RTC when both drugs were added at the beginning of the experiment (Fig. 4). The better activity ofboth drugs in combination, at least within 3 h of incubation, than either component alone was less obvious when hypertonic osmotically stabilized brain heart infusion broth instead of Mueller-Hinton broth was used for the growth curves (Fig. 5). Addi- tion ofRTC in a subinhibitory concentration 3 h h after pretreatment with penicillin, however, FIG. 3. Timed killing curves ofE. coli by penicil- significantly decreased optical density, sug- lin or RTC added 2 h after pretreatment (arrow) of gesting bactericidal activity of RTC. At this the bacteria with bacteriostatic RTC or bacterio- time 99.9% of all bacteria were transformed to static/bactericidal penicillin concentrations. 806 DASCHNER ANTIMICROB. AGENTS CHEMOTHER. penicillin alone or in combination with chloramphenicol. Some recent experimental studies also question the concept ofgeneral antagonism between bacteriostatic and bactericidal drugs. 0086 Luboshitzky et al. (16) found striking examples of synergism between tetracyclines/cephalori- 007 dine on Serratia and between tetracycline/kan- amycin on Enterobacter. The combination of chloramphenicol and cephalothin resulted in an additive effect against four out of six multiple- antibiotic-resistant strains of Staphylococcus epidermidis (24). None of these studies, how- 05.0

optical density

QO03

* control 1irlOpg /ml PEN 0.02 0.02 /* 0 0.90pg / ml RTC 0.90 RTC + 3h P EN * pg/mI aftwr IlSOpg /ml PEN+ Q01 t * 1 50gg/ml OSO0g/ml 150pg/ml PEN+0°90t"9/ml RTC

L ~~~~~~~~~~~~h 1 2 3 4 5 6 7 FIG. 4. Growth ofE. coli at subinhibitory concentrations ofpenicillin orRTC alone or in combination in Mueller-Hinton broth. Addition of penicillin or RTC both in subinhibitory concentrations to cells pretreated with the other drug led to prompt bacteriostasis. danger of severe side effects. It is, however, generally believed that, using two drugs, bactericidal and bacteriostatic antibiotics should not be combined because of their in vitro and in vivo antagonism. Antibiotic antagonism, especially between the penicillins chloramphenicol and tetracycline (13, 17, 18, 23), could be demonstrated in various clinical studies. Similarly designed studies, however, failed to show any untoward effect of combinations of bacterio- FIG. 5. Growth ofE. coli at subinhibitory penicil- static and bactericidal drugs as compared with lin or RTC concentrations alone or in combination in using only one drug. Finland et al. (9) ob- osmotically stabilized brain heart infusion broth. served no difference in the antibacterial activ- Addition ofRTC in subinhibitory concentrations to ity of human plasma when either penicillin or penicillin-induced spheroplasts resulted in bacteri- penicillin combined with chloramphenicol was cidal activity (arrow). administered to human subjects. Ahern and Kirby (1) treated 25 patients with pneumococ- TABLE 2. Susceptibility (MIC) offour staphylococcal cal pneumonia with penicillin alone and 25 L-form and parent strains to RTC with penicillin and chlortetracycline and found MIC (,Ug/m1) neither antagonism nor a beneficial effect from Strain Parent form L-form the drug combination. Walker (25) treated patients with streptococcal pharyngitis with peni- 36 125 1.9 cillin alone or with penicillin and chloramphen- 65 500 3.9. icol; duration of fever, frequency of relapses, or 68 1.9 0.9 101 125 0.45 other symptoms were not influenced by using VOL. 10, 1976 COMBINATION OF BACTERIOSTATIC/CIDAL DRUGS 807 ever, considered the original finding of Jawetz where tetracyclines were shown to be effective (10) that results of drug combinations may de- in vitro (3). Further clinical studies are neces- pend on the bacterial strains, drug concentra- sary to prove the effectiveness of these drug tions, time of interaction, and other laboratory combinations. One of these studies is currently variables. I therefore investigated some of being done in our hospital. At present, how- these variables and found not antagonistic, but ever, I would not recommend the use of combi- additive, indifferent, or synergistic activity be- nations of bactericidal and bacteriostatic drugs tween cephalothin/penicillin and tetracycline unless antagonism has been excluded, prefera- on various strains ofS. aureus andE. coli when bly by methods using bactericidal end points. tested by using the checkerboard dilution technique. Slight antagonism between penicillin/ ACKNOWLEDGMENTS cephalothin and RTC, each combined in bacte- The enthusiastic technical assistance of Ingrid Emricc- ricidal and bacteriostatic concentrations, oc- son is appreciated. curred only between 3 and 8 h, but not at 24 h, This work was supported in part by a grant from Hoechst of incubation. There was also no antagonism AG, Frankfurt, Germany. when the drugs were combined in bacteriostatic concentrations. This was also true if penicillin/ LITERATURE CITED cephalothin or RTC was added to E. coli cells 1. Ahern, J. J., and W. M. M. Kirby. 1953. Lack of inter- pretreated with either bacteriostatic RTC or ference of aureomycin with penicillin in treatment of penicillin concentrations. Penicillin as well as pneumococcic pneumonia. Arch. Intern. Med. 91:197- cephalothin could exert a bactericidal effect 203. 2. Brumfitt, W. 1971. Antibiotic combinations. Lancet after a 2-h exposure ofE. coli to bacteriostatic ii:387-389. RTC concentrations. The latter findings may 3. Chow, A. W., V. Patten, and L. B. Guze. 1975. Compar- account for some of the conflicting results of in ative susceptibility of anaerobic bacteria to minocy- vitro and in vivo studies concerning the effect of cline, doxycycline, and tetracycline. Antimicrob. Agents Chemother. 7:46-49. various drug combinations, especially those be- 4. Daschner, F. 1975. Synergismus von Cefazolin und To- tween so-called primarily bacteriostatic and bramycin bei Klebsiella pneumoniae als Erreger no- bactericidal antibiotics (1, 9, 14). Even the com- sokomialer Infektionen. Infection 3(Suppl. 1):31-34. bined action of subinhibitory penicillin and 5. Daschner, F. 1976. Inhibition of cell wall synthesis by sulfonamides and trimethoprim. Chemotherapy RTC concentrations invariably resulted in 22:12-18. more pronounced inhibition of growth than did 6. Daschner, F. D., R. Hovel, and W. Marget. 1973. Ac- using either drug alone. Bactericidal activity of tion of gentamicin on penicillin G-induced Proteus subinhibitory RTC concentrations on penicil- spherosplasts. Chemotherapy 18:235-241. 7. Dimmling, T. 1972. Vergleichsstudie uber drei derzeit lin-induced E. coli spheroplasts in osmotically gebrauchliche Tetracycinederivate zur intraven6sen stabilized fluid medium suggested that cell Anwendung (Serum- und Urinspiegel). Med. Klin. wall-defective bacteria (spheroplasts) might 67:1632-1637. permit greater access ofRTC into the bacterial 8. Dowling, H. F. 1960. The conditions under which the use of combinations of antibiotics is undesirable. An- cell, which thus better reaches its site ofaction. tibiot. Chemother. 8:114-125. This hypothesis could be supported by compar- 9. Finland, M., C. V., Pryles, and W. F. Jones. 1958. ing the RTC MICs of staphylococcal L-forms Antibiotic combinations. Antibacterial action of and their corresponding parent forms with in- plasma of human subjects after ingestion of penicillin V or chloramphenicol or both. N. Engl. J. Med. tact cell walls. L-forms were invariably more 258:817-824. susceptible to RTC than the parent forms. 10. Jawetz, E. 1968. Combined antibiotic action: some defi- My data confirm those of Mouton and Koel- nitions and correlations between laboratory and clini- man (19) and Jawetz (10), who found that cal results, p. 203-209. Antimicrob. Agents Chemo- drug ther. 1967. interaction patterns between bactericidal and 11. Jawetz, E., and J. B. Gunnison. 1952. Studies on anti- bacteriostatic drugs are highly dependent on biotic synergism and antagonism: a scheme of com- bacterial species, strains, and drug concentra- bined antibiotic action. Antibiot. Chemother. 2:243- tions used. Better activity of tetracyclines on 248. 12. Kagan, B. M. 1968. Antibiotic sensitivities of staphylo- cell wall-defective bacteria as compared with coccal L-forms, p. 314-318. In L. B. Guze (ed.), Micro- bacterial cells with intact cell walls has been bial protoplasts, spheroplasts and L-forms. The Wil- demonstrated previously by Kagan (12). liams & Wilkins Co., Baltimore. Whether this may account, at least in part, for 13. Lepper, M. H., and H. F. Dowling. 1951. Treatment of pneumococcal meningitis with penicillin compared the additive or synergistic action ofpenicillin in with penicillin plus aureomycin. Arch. Intern. Med. combination with RTC on certain strains re- 88:489-494. mains to be studied. Combinations of bacteri- 14. Lepper, M. H., G. G. Jackson, H. F. Dowling, and J. cidal drugs with tetracyclines may be necessary Seto. 1953. Comparison of results of treatment of pneumococcic infection in mice with multiple doses of to broaden the spectrum in the treatment of aureomycin and penicillin combined with each drug certain mixed infections, e.g., with anaerobes, alone. Am. J. Med. Sci. 225:648-656. 808 DASCHNER ANTIMICROB. AGENTS CHEMOTHER. 15. Limson, M. B., and R. B. Guevara, Jr. 1962. Efficacy of 20. Russell, F. E. 1962. Synergism between sulfonamide a single intravenous injection ofpyrrolidinomethylte- drugs and antibiotics of the polymyxin group against tracycine in pneumococcal pneumonia: report of 20 Proteus species in vitro. J. Clin. Pathol. 16:362-366. cases with assays of 24 hour antibiotic activity in 21. Sabath, L. D. 1968. Synergy of antibacterial substances serum. Antibiot. Chemother. 12:85-90. by apparently known mechanisms, p. 210-217. An- 16. Luboshitzky, R., T. Sacks, and J. Michel. 1973. Bacte- timicrob. Agents Chemother. 1967. ricidal effect of combinations of antibiotics on Klebsi- 22. Simmons, N. A. 1969. Potentiation ofinhibitory activity ella-Enterobacter-Serratia. Chemotherapy 19:354- of colistin on Pseudomonas aeruginosa by sulfame- 366. thoxazole and sulfamethiazole. Br. Med. J. 3:693- 17. McCabe, W. R., and G. G. Jackson. 1965. The treat- 696. ment of pyelonephritis. Bacterial, drug and host fac- 23. Strohm, J. 1955. The question of antagonism between tors in success or failure among 252 patients. N. Engl. penicillin and chlortetracycine, illustrated by thera- J. Med. 272:1037-1044. peutical experiments in scarlatina. Antibiot. Med. 18. Mathies, A. W., Jr., J. M. Leedom, D. Ivler, P. F. 1:6-12. Wehrle, and B. Portnoy. 1968. Antibiotic antagonism 24. Traub, W. H. 1970. In vitro activity of chloramphenicol in bacterial meningitis, p. 218-224. Antimicrob. combined with cephalothin against DNase-positive, Agents Chemother. 1967. multiple-antibiotic resistant strains of Staphylococ- 19. Mouton, R. P., and A. Koelman. 1966. The interaction cus epidermidis. Chemotherapy 15:234-241. patterns of combined antibacterial agents. Experi- 25. Walker, S. H. 1953. A study of the possible interference ments with two agar diffusion methods. Chemother- of chloramphenicol with penicillin in acute strepto- apy 11:10-26. coccal pharyngitis. Antibiot. Chemother. 3:677-680.