ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, July 1974, p. 93-97 Copyright 0 1974 American Society for Microbiology Vol. 6, No. 1 Printe in U.S.A. Lack of Activity of Sulfamethoxazole an Trimethoprim Against Anaerobic Bacteria J. E. ROSENBLATT AND P. R. STEWART Meical Service, Wasworth Veterans Aministration Hospital Center, Los Angeles, California 90073, an Department of Meicine, UCLA School of Meicine, Los Angeles, California 90024 Receive for publication 15 April 1974 The activity of sulfamethoxazole (SMX), trimethoprim (TMP), an the combination of the two was etermine against a variety of anaerobic bacteria. Brucella agar was somewhat inhibitory for SMX an TMP but activity was goo an equivalent in Diagnostic Sensitivity Test Agar (Oxoi) an Mueller-Hinton agar an the latter was selecte for use in these stuies. Agar ilution susceptibility tests showe that 95 of 98 anaerobic isolates were resistant to > 100,ug of SMX per ml an 85 were resistant to > 6.25 jtg of TMP per ml. "Checkerboar" agar ilution stuies of combine activity showe that 66 of 72 isolates were resistant to > (100 Ag of SMX per ml + 6.25,ug of TMP per ml) an only six isolates were susceptible to the synergistic activity of the combination. The majority of 32 isolates teste by the isk iffusion metho were also resistant to SMX an TMP iniviually an to the combination 25-,ug isk. Correlation between agar ilution minimal inhibitory concentration an isk zone size results was in general goo for iniviual agents. Four Bacteroies fragilis isolates were inhibite by the combination 25-ltg isk but were resistant to SMX + TMP by agar ilution "checkerboar." This iscrepancy may have been ue to ifferent incubation perios since isk results also showe resistance when rea after 48 h (as is one with agar ilution) rather than the stanar 24 h for isk tests. These stuies suggest that SMX an TMP, either iniviually or in combination, are not active against the great majority of anaerobic bacteria. Co-trimoxazole is an antimicrobial agent recently introuce into the Unite States for the treatment of chronic urinary tract infections ue to susceptible organisms, primarily gramnegative bacilli. This agent has also been use for some time in Englan an Europe for the treatment of pulmonary infections (4), typhoi fever (2), an gonorrhea (9). Co-trimoxazole is a fixe ratio combination of sulfamethoxazole (SMX) an trimethoprim (TMP). Antimicrobial activity epens upon synergism between SMX an TMP which both interfere with bacterial folic aci activity in the synthesis of nucleic acis, though at ifferent sites (7). We ecie to investigate the potential for use of co-trimoxazole in the treatment of anaerobic infections by stuying the activity of SMX an TMP iniviually an in combination against a variety of anaerobic bacteria. Since the in vitro activity of SMX an TMP is known to be meia epenent (12) an since growth of anerobes may vary on ifferent meia we initially ha to etermine the most appropriate meium for these susceptibility stuies. Subse- 93 quently, we use agar ilution an isk iffusion methos to etermine the activity of SMX an TMP against anaerobes. MATERIALS AND METHODS Antimicrobials an bacteria. TMP lactate was provie as stanar laboratory power by Burroughs Wellcome an Co., Inc. SMX stanar power was obtaine from the Hoffmann-La Roche Co., Inc. Stock solutions of each antimicrobial were prepare in concentrations 10 to 20 times the esire final concentration by using sterile istille water as the iluent. Aition of a few rops of 1 N NaOH was necessary to issolve the SMX. Portions of each antimicrobial were ispense an frozen. Iniviual TMP (1.25,g) an SMX (23.75 Mg) isks were supplie by Burroughs Wellcome an Co. Combine SMX + TMP (25 Ag) isks were manufacture by Difco. The 98 anaerobe strains teste were stock cultures of the Wasworth Anaerobe Laboratory. Seventythree were clinical isolates an 25 were from the normal human fecal flora. Meia an proceures. Thioglycolate broth (without inicator), brucella agar (BA), an brucella broth were BBL proucts an Mueller-Hinton agar (MHA)
94 ROSENBLATT AND STEWART an broth were Difco proucts. Diagnostic sensitivity test agar (DST) an broth were manufacture by Oxoi Lt. Fresh efibrinate horse bloo was lyse by alternate freezing an thawing. Stock solutions of vitamin 100,000 gsg/ml, Kl, (Nutritional Biochemicals Corp.) were prepare in ethanol an refrigerate. MHA an DST were supplemente with 5% lyse horse bloo an vitamin K1 (10,g/ml), whereas BA was supplemente with 5% sheep bloo an vitamin K, (10 zg/ml). Agar ilution (Steer's replicator [10 1) an isk iffusion susceptibility tests were performe by using methos escribe by Sutter et al. (11). Stuies of combine SMX + TMP activity were one by the "checkerboar" metho by using fixe an varying concentrations of each antimicrobial as escribe by Sabath (8). SMX in concentrations of 200, 100, 50, an 25 ug/ml was combine with TMP in concentrations of 12.5, 6.25, 3.12, an 1.56 /g/ml. For example, combination plates containe TMP (12.5,Ag/ml) an SMX (200,g/ml), TMP (12.5 /Ag/ml) an SMX (100,ug/ml), an TMP (12.5 Ag/ml) an SMX (50,g/ml), etc. Disk zone of inhibition iameters were measure by using vernier calipers an agar ilution minimum inhibitory concentrations (MIC) were rea as the lowest concentration showing complete inhibition of growth. In a few instances the MIC was rea as that concentration allowing growth of three or four colonies when this represente a rastic inhibition of growth compare to the next lower concentration. Anaerobic incubation was carrie out by using the GasPak system (BBL). Agar ilution plates were rea after 48 h of incubation, an isk plates were rea at 24 h (ata reporte in Results) an again at 48 h. RESULTS Anaerobes stuie. Table 1 shows the number an kins of anaerobic isolates stuie. The susceptibility of 98 ifferent isolates to SMX an TMP iniviually was etermine by the agar ilution metho. Seventy-two of these were inclue in "checkerboar" stuies of combine SMX + TMP activity. A total of 32 isolates was stuie by the isk iffusion metho an 10 of these were Bacteroies fragilis. Determination of appropriate meium. Initially, the activity of SMX an TMP in BA an MHA was stuie by comparing the susceptibility of all 98 isolates in both meia by using the agar ilution metho. A ifference of at least two ilutions in the MIC was consiere evience of a significant ifference in antimicrobial activity in the two meia. Table 2 shows that SMX was less active in BA than in MHA against 6 of 98 isolates an that TMP was less active in BA against 10 isolates. On the other han, neither SMX nor TMP were less active in MHA than BA against any of the anaerobes. On the basis of these ata suggesting ecrease TABLE 1. Anaerobes stuie for susceptibility to SMX an TMP Agar ilution Disk iffusion Bacteroiesfragilis 38a (26)" 1oc Bacteroies melaninogenicus 4 (4) 4 Bacteroies oralis 4 (3) 1 Fusobacterium mortiferum 4 (3) 2 Fusobacterium necrophorum 4 (3) 1 Fusobacterium nucleatum 2 (1) 1 Fusobacterium varium 4 (3) 1 Clostriium perfringens 8 (6) 3 Clostriium ramosum 4 (3) 2 Eubacterium limosum 4 (3) 1 Peptococci 7 (6) 1 Peptostreptococci 12 (8) 5 Propionibacterium acnes 3 (3) 0 Total 98 (72) 32 'Number of isolates stuie for susceptibility to SMX an TMP iniviually. b Number of isolates stuie by the "checkerboar" metho for susceptibility to combine activity of SMX + TMP. 'Number of isolates stuie for susceptibility to iniviual an combine activity of SMX an TMP by isk iffusion metho. TABLE 2. ANTIMICROB. AG. CHEMOTHER. Differences in activity of SMX an TMP in ifferent meiaa Antimicrobial act in: SMX TMP BA" < MHAC 6 10 MHA < BA oe 0 a Activity stuie by agar ilution etermination of MIC against 98 anaerobic isolates. Brucella agar. cmueller-hinton agar. Number of isolates against which SMX (or TMP) was less active in BA than in MHA. e Number of isolates against which SMX (or TMP) was less active in MHA than in BA. SMX an TMP activity in BA, this meium was eliminate from the stuy. Similar agar ilution stuies were carrie out to compare the activity of SMX an TMP in MHA an DST against 20 anaerobic isolates (Table 3). SMX an TMP were less active in DST than in MHA against one an two isolates, respectively. On the other han, in no instance was either agent less active in MHA than in DST. On the basis of these ata, we etermine that there was no significant ifference in antimicrobial activity between these two meia, an selecte MHA as our stanar meium for
VOL. 6, 1974 SULFAMETHOXAZOLE AND TRIMETHOPRIM AGAINST BACTERIA 95 testing susceptibility of anaerobes to SMX an TMP. All subsequently escribe stuies were carrie out by using MHA. Susceptibility of anaerobes to SMX an TMP. Table 4 shows the susceptibility of 98 anaerobic isolates to SMX an TMP, teste by the agar ilution metho. Only three isolates ha an MIC of SMX < 25,ug/ml an woul be consiere susceptible. The other 95 isolates were resistant to > 100 Ag/ml. Likewise, only 13 isolates were susceptible to < 3.12 ug of TMP per ml (susceptible), whereas 85 were resistant to > 6.25 Mg/ml. These ata inicate that the great majority of anaerobes stuie were resistant to SMX an TMP as iniviual agents. The susceptibility of 72 anaerobic isolates to the combine activity of SMX + TMP as teste by the agar ilution "checkerboar" metho is shown in Table 5. There was no emonstrable synergistic activity between the two agents against 66 isolates. Only six isolates (two Clostriium ramosum, two Peptostreptococcus intermeius, an two Bacteroies melaninogenicus) were susceptible to SMX + TMP synergy by Sabath's criteria (8). These were inhibite by TABLE 3. Differences in activity of SMX an TMP in ifferent meiaa Antimicrobial act in: SMX TMP DSTb < MHAC l 2 MHA < DST oe 0 aactivity stuie by agar ilution etermination of MIC against 20 anaerobic isolates. ' Oxoi iagnostic sensitivity test agar. cmueller-hinton agar. Number of isolates against which SMX (or TMP) was less active in DST than in MHA. e Number of isolates against which SMX (or TMP) was less active in MHA than in DST. TABLE 4. Susceptibility of 98 anaerobic isolates to SMX an TMPa Determination SMX TMP MIC5 (,ug/ml) <25 > 100 < 3.12 6.25 No. isolates 3c 95 13e 85' a Agar ilution metho using MHA. 'Minimum inhibitory concentration. cnumber of isolates consiere susceptible to SMX. Number of isolates consiere resistant to SMX. e Number of isolates consiere susceptible to TMP. ' Number of isolates consiere resistant to TMP. TABLE 5. Susceptibility of 72 anaerobic isolates to the combine activity of SMX + TMPa Determination Synergism No synergism MICb (1g/Ml) < {SMX 25 > SMX 100 MICb(~gml) ktmp" 1.56 ~ TMP 6.25 No. isolates 6c 66 a Agar ilution "checkerboar" metho using MHA. bminimum inhibitory concentration. c Number of isolates against which SMX + TMP showe synergistic activity. Number of isolates against which SMX + TMP i not show synergistic activity. concentrations of SMX an TMP (in combination) which woul be consiere therapeutically achievable an which approximate the 20: 1 concentration ratio thought to be optimum for synergism (25 Ag of SMX per ml an 1.56,ug of TMP per ml). The 66 resistant isolates were inhibite only by concentrations of SMX an TMP > (100,g/ml an 6.25,g/ml), respectively. The results of isk iffusion susceptibility tests with 32 isolates are shown in Table 6. There are no zone size stanars for interpretation of isk tests with SMX, TMP, an the combination. However, 24, 28, an 16 isolates ha no inhibition zones when teste with SMX (23.75,Ag), TMP (1.25 Mg), an combination (25 Mug) isks, respectively. These isolates woul certainly be consiere resistant. Eight, four, an five isolates exhibite inhibition zones > 13 mm, with SMX, TMP, an combination isks, respectively. These isolates might be consiere susceptible or "moerately resistant" in some instances. There were 11 other isolates with zones aroun the combination isk but these coul be attribute to inhibitory activity of one of the iniviual component agents rather than synergism of the combination. It shoul be note that although the term "synergism" is use to enote enhance activity with the combination isk, in some instances this activity may be ue to aitive antimicrobial activity rather than true synergism. Correlation between MIC an zone sizes for iniviual agents was generally goo. The exceptions were three isolates (two peptostreptococci an a B. melaninogenicus) with SMX MIC of > 100 an SMX isk zones of 32.3 to 47.0 mm. There is no reay explanation for this iscrepancy. Perhaps it was relate to poor growth of these isolates on the isk susceptibility plates (although there were no zones aroun TMP isks in the two isolates with high TMP
96 ROSENBLATT AND STEWART MIC) or variability in the interpretation of SMX inhibition of growth on the agar ilution or isk plates or both. Table 7 shows the correlation between MIC an isk zones in those five isolates showing susceptibility to a synergistic effect of SMX + TMP; i.e., the combine isk ha an inhibition zone, whereas the iniviual isks i not. All four B. fragilis isolates faile to show susceptibility to SMX + TMP synergism by the agar ilution metho. Only against the single B. melaninogenicus TABLE 6. Susceptibility of 32 anaerobic isolates to SMX, TMP, an the combination by the isk iffusion metho Zone of inhibition SMX m TM 12Ag +SMX iameter (23.75 sg) TMP (125 pg) (25TMg) (mm) (5pg 6h 24C 28 16 > 13 8 4 5 a Disk content of each antimicrobial. 'No zone of inhibition present since isk iameter is 6 mm. c Number of isolates in each zone size category. Eleven other isolates ha zone sizes in this category which were attributable to the iniviual activity of SMX or TMP rather than synergism of the combination. TABLE 7. isolate i the SMX + TMP combination show synergism by both methos. There were six other B. fragilis isolates which were totally resistant by the isk metho an all B. fragilis isolates were totally resistant by agar ilution. An explanation for the iscrepancy with isk testing of B. fragilis may be foun in the time of reaing of results. Disk susceptiblity test results were rea (an reporte here) after 24 h of incubation an in most cases this correlate well with a subsequent 48-h reaing. However, with the four B. fragilis isolates liste in Table 7, the zone aroun the combine SMX + TMP isk was markely reuce at 48 h. In fact, in three of the four isolates there was no zone at 48 h. Therefore, the 48-h isk test reaing appears to correlate best with results obtaine by agar ilution (rea at 48 h), which inicate a lack of any synergistic activity of SMX + TMP against B. fragilis. In spite of these possible iscrepancies in results, the great majority of anaerobes, when teste by either the agar ilution or isk iffusion methos, were resistant to SMX an TMP iniviually an to the combination. No specific anaerobe was consistently susceptible. DISCUSSION This stuy emonstrates that neither SMX nor TMP, iniviually or in combination, have Comparative susceptibility offive anaerobic isolates to SMX, TMP, an the combination by the agar ilution0 an isk iffusion methos MIC" (pig/ml) Disk zone (mm) Anaerobe isolates m x TM x AnaerobeisolatsSMX TMP + TMP 23.75 pie 1.25 pg + TMP 25~~~~~~~~~~~5A Bacteroies melaninogenicus > 100 > 6.25 25 6e 6 19.5' Bacteroies fragilis #1 > 100 > 6.25 > 100 6 6 18.5 >6.25 661. Bacteroies fragilis #2 > 100 >6.6.2 > 100 6 6 15.3 Bacteroies fragilis #3 > 100 > 6.25 100 6 6 18.5 3.12 Bacteroiesfragilis#4 >100 >6.25 >100 6 6 13.5 >6.25 661. 0"Checkerboar" metho using MHA. 'Minimum inhibitory concentration (results rea after 48 h of incubation). 'Content of antimicrobial in each isk. Combine MIC of <25 pg of SMX per ml an 1.56 jig of TMP per ml inicates synergism an > 100 pg of SMX per ml an 6.25 pg of TMP per ml inicates no synergism. 'No zone of inhibition present since isk iameter is 6 mm. I Zone sizes when rea after 24 h of incubation. ANTIMICROB. AG. CHEMOTHER.
VOL. 6, 1974 SULFAMETHOXAZOLE AND TRIMETHOPRIM AGAINST BACTERIA 97 significant antimicrobial activity against a variety of anaerobic bacteria. These finings are similar to those of Bushby (1) who coul not emonstrate increase susceptibility of three strains of Bacteroies (unspeciate) to the SMX + TMP combination. These strains were reporte to be resistant to TMP but susceptible to SMX. Naffs (5) emonstration that feeing of co-trimoxazole to human volunteers eliminate Enterobacteriaceae from their fecal flora but i not affect Bacteroies also suggests lack of activity against these anaerobes. On the other han, Okubaejo et al. (6) reporte inhibition of more than 100 Bacteroies strains by 25-Ag co-trimoxazole isks an inhibition of all 60 B. fragilis strains teste by an agar ilution metho. This report, however, is in the form of a "letter-to-the-eitor" an minimal information is provie. The authors o state use of DST an anaerobic incubation in an atmosphere containing 95% H2 an 5% CO2. The iscrepancies between these results an our own are not easily explaine. We emonstrate that results in DST an MHA (use in our stuies) shoul be similar an our anaerobic incubation (Gas- Pak system) shoul have provie a comparable atmosphere. Supplementation of our meia with vitamin K1, use of a stanar heavy inoculum, an incubation of agar ilution plates for 48 h may have prouce heavier anaerobic growth than that obtaine by Okubaejo et al. This in turn woul ten to result in iminishe antimicrobial activity. Our methos are similar to those use by Sutter et al. (11) in extensive investigations of the susceptibility of anaerobes to various antimicrobials. These stuies have further emonstrate that BA, a meium recommene for anaerobe susceptibility testing, is not suitable for use with SMX an TMP because of inhibitory activity. Demonstration of aequate anaerobic growth on MHA (containing 10 ug of vitamin K1 per ml) allowe us to use this meium, which when supplemente with 5% lyse horse bloo is not inhibitory for SMX an TMP. Koch an Burchall (3) have recently shown that high concentrations of thymiine in commercially prepare meial can reverse the antimicrobial activity of TMP. Although BA was not inclue in that stuy, a similar enriche meia which allows goo anaerobic growth (brain heart infusion) containe as much as 30.9 ug of thymiine per ml, whereas DST an MH broth formulations both containe less than 1.0 Mg of thymiine per ml. Excess thymiine content may be one of the factors contributing to BA inhibitory activity against SMX an TMP. In conclusion, our stuies inicate that the majority of anaerobes are not susceptible to SMX, TMP, or the combination, an suggest that co-trimoxazole will not be useful for treatment of anaerobic infections. There is at least one other conflicting report inicating susceptibility of B. fragilis- to co-trimoxazole. However, this agent shoul not be use to treat anaerobic infections until further experience with in vitro stuies or experimental infections provies information which will help to resolve these questions. LITERATURE CITED 1. Bushby, S. R. 1973. Trimethoprim-sulfamethoxazole: in vitro microbiological aspects. J. Infect. Dis. 128 (Suppl. 1):S442-S462. 2. Kamat, S. A. 1970. Evaluation of therapeutic efficacy of trimethoprim-sulphamethoxazole an chloramphenicol in enteric fever. Brit. Me. J. 3:320-322. 3. Koch, A. E., an J. J. Burchall. 1971. Reversal of the antimicrobial activity of trimethoprim by thymiine in commercially prepare meia. Appl. Microbiol. 22:812-817. 4. Lal, S., an K. K. Bhalla. 1969. Comparison of tetracycline an trimethoprim-sulphamethoxazole in acute episoes in chronic chest infections. Postgra. Me. J. 45(Suppl.) :91-94. 5. Naff, H. 1971. On the changes in the intestinal flora inuce in man by Bactrim'i. Pathol. Microbiol. 37:1-22. 6. Okubaejo, 0. A., P. J. Green, an D. J. Payne. 1973. Bacteroies in the bloo. Lancet 1:147. 7. Reeves, D. S. 1971. Sulphamethoxazole/trimethoprim: the first two years. J. Clin. Pathol. 24:430-437. 8. Sabath, L. D. 1968. Synergy of antibacterial substances by apparently known mechanisms, p. 210-217. Antimicrob. Ag. Chemother. 1967. 9. Schofiel, C. B., G. Masterton, M. Moffett, an M. I. McGill. 1971. Gonorrhea in women: treatment with sulfamethoxazole an trimethoprim. J. Infect. Dis. 124:533-538. 10. Steers, E., E. L. Foltz, an B. S. Graves. 1959. An inocula replicating apparatus for routine testing of bacterial susceptibility to antibiotics. Antibiot. Chemother. 9:307-311. 11. Sutter, V. L., H. R. Attebery, J. E. Rosenblatt, K. S. Bricknell, an S. M. Finegol. 1972. Anaerobic bacteriology manual. Department of continuing eucation in health sciences, university extension an the school of meicine, UCLA. 12. Waterworth, P. M. 1969. Practical aspects of testing sensitivity to trimethoprim an sulphonamie. Postgra. Me. J. 45(Suppl.):21-27.