748
New Antimicrobial Agents
In Vitro Susceptibility of
Haemophilus influenzaeto Cefaclor, Cefixime, Cefetamet and Loracarbef J.D. Williams*, M. Powell, Y.S. Fah, A. Seymour, M. Yuan
The susceptibility of 2,212 Haemophilus influenzae isolates cultured in UK clinical laboratories in 1991 was determined for four orally-administered [3-1actam drugs. These isolates included 1,893 ampicillin-susceptible, 191 I~-lactamase-positive and 128 ampiciilin-resistant, [3-1actamase-negative Haemophilus influenzae. While 150 (6.8 %) isolates were resistant to cefaclor (MIC > 16 rag/I) and 85 (3.8 %) to loracarbef, all were inhibited by < 2 mg/! cefetamet and _ 0.5 mg/l cefixime showed this type of resistance to ampicillin. Results indicate the importance of detecting non[3-1actamase-mediated resistance to ampicillin and any concomitant diminished susceptibility to other 13-1actam drugs.
In 1990 a UK survey of sputum isolates showed that the prevalence among 13-1actamase-positive and I~-lactamase-negative isolates of resistance to ampicillin (9.4 % and 5.2 % respectively) (1) had increased since the last survey of England and Scotland (6.2 % and 4 %) conducted in 1986 (2). The higher percentage of l~-lactamase-positive isolates noted was similar to figures obtained in Departmentof MedicalMicrobiology,The LondonHospital Medical College,Turner Street, London E1 2AD, UK.
Vol. 11. No. 8
Wales and Ireland of 8.4 % and 10.2 % respectively (3, 4). However, non-I~-lactamase-mediated resistance to ampicillin was not detected by UK laboratories in 85 of 100 such isolates collected during the 1986 survey. This is of importance because the mechanisms involved affect not only susceptibility to ampicillin but also commonly produce a marked degree of diminished susceptibility to most penicillins and cephalosporins with corresponding therapeutic implications. For example, the reduction in susceptibility to cefaclor and cefixime shown by such organisms was reported for the 100 13-1actamase-negative isolates obtained in the 1986 survey, which required 1-64 mg/l ampicillin for inhibition. While only 2 % of the total number of organisms collected (2458) were resistant to cefaclor, 47 of 50 cephalexin-resistant isolates were ampicillinresistant and ~-lactamase negative, as were all four isolates which were resistant to ceflxime. The modal MICs of the two cephalosporins for organisms with this mechanism of resistance were two to three doubling dilutions above modal values for the ampicillin-susceptible and 13-1actamase-positive groups. In the present study the 2,212 clinical isolates of
Haemophilus influenzae were examined for their susceptibility to three orally administered cephalosporins (cefaclor, cefixime and cefetamet) and to the carbacephem loracarbef. Susceptibility to these 15-1actam drugs was compared between groups of Haemophilus influenzae delineated according to their susceptibility to ampicillin.
Materials and Methods. All media, supplements and discs were obtained from Unipath, UK. Ampicillin was obtained from Sigma Chemical, USA, amoxicillin-clavulanate (in the form of'Adatabs') from Mast Laboratories, UK, cefaclor and loracarbef from Lilly Industries, UK, cefetamet from Roche Products, U K and cefixime from Ledefle Laboratories, UK. Methods were those used in previous studies (2). Briefly, five-hour cultures of organisms in brain heart infusion broth supplemented with 5 % (vol/vol) Fildes' supplement were diluted (1/100) in peptone water. The suspension was used for swab inoculation of several DST agar plates, supplemented with 0.25 % (vol/vol) lysed horse blood and 10 mg/1 NAD, to which the following discs were applied: cefaclor 30 lag, loracarbef 30 gg, cefetamet 10 gg and cefixime 5 ~g. The same suspension provided an inoculum of 104 cfu/ spot when deposited by a multipoint inoculator onto similar agar containing antimicrobial agents
Vol. 11, 1992
749
at final concentrations ranging from 256 to 0.008 mg/l. Incubation was for 18 h at 37 °C in an atmosp h e r e of 5 % CO2 and 95 % air. Control strains Haemophilus influenzae A T C C 491002 (fully-susceptible), Haemophilus influenzae W10311 ([3lactamase-positive, chloramphenicol and tetracycline-resistant) and Haemophilus influenzae 518 (ampicillin-resistant, 13-1actamase-negative) were included in all batches tested. Production of I~-lactamase was detected by the iodometric m e t h o d (5).
Results and Discussion. Using the same inhibition zone diameter and MIC breakpoints adopted in previous studies and those suggested by manufacturers in the case of cefetamet and loracarbef, the numbers o f isolates resistant to the agents tested are shown in Table 1. The susceptibility of all isolates to the three cephalosporins and to loracarbef is c o m p a r e d in Table 2, in which MIC50 and MIC90 values are given for organisms grouped according to their susceptibility to am-
Table 1:
picillin. As expected, a rise in b o t h values of two to three doubling dilutions was seen for three drugs between the ampicillin-susceptible group and those organisms which showed non-l~-lactamase-mediated diminished susceptibility to ampicillin. T h e exception was cefixime although the reason for this was partly the fact that the distribution of all isolates according to the M I C of cefixime was generally in a broader, flatter curve compared with other 15-1actam drugs. Table 3 shows similar grouping of isolates according to inhibition zone diameter, comparing range and modal values for the four l~-lactam antimicrobial agents. While there was little discernible difference in modal values between the ampicillin-susceptible and 13-1actamase-positive groups, the range extended downwards for the latter group except in the case o f cefixime. For all four agents these two parameters shifted downwards in the I~-lactamase-negative, ampicillinresistant group.
Preva~ence~fresistancet~ampici~inandf~ur~therr3-~actamagentsam~ng2~2~2Haem~philusin~uenzaeis~ates.
Antimicrobial agent
Criteria for resistance MIC (mg/l)
Ampieillin Ampicillin Cefaelor Loracarbef Cefetamet Cefixime
Number (%) resistant 1991
not applicable >1 _>16 > 16 >8 >2
191a 128b 150 85 none none
(8.6) (5.8) (6.8) (3.8) (0) (0)
aBeta-lactamase positive isolates. bBeta-laetamase negative isolates.
Table 2: Comparison of the susceptibility of 2,212 Haemophilusinfluenzaeisolates to ampicillin and to four other I%lactam antimicrobial agents. MIC (mgll) Susceptibility to ampieillin
Cefaclor
Loracarbef
Cefetamet
Cefixime
MIC50
MIC90
MIC50
MIC90
MIC50
MIC90
MIC50
MIC90
Ampicillin-susceptible (n=1893)
4
8
2
4
0.12
0.25
0.06
0.12
Ampicillin-resistant beta-lactamase positive
4
8
2
8
0.25
0.5
0.03
0.06
16
64
8
32
0.5
1
0.06
0.25
(n =
191)
Ampicillin-resistant beta-taetamase negative (n = 128)
750
Eur. J. Clin. Microbiol. Infect. Dis.
Table 3: Comparison of the susceptibility of 2,212 Haemophilus influenzae isolates to ampiciUin and to four other 13-1attain
antimierobial agents according to inhibition zone diameters. III II II IIIIIII IIIII II
hibition zone diameters (mm) Susceptibility to ampieillin
Cefaclor
Loraearbef
Cefetamet
Cefixime
Range
Mode
Range
Mode
Range
Mode
Range
Mode
Ampi¢illin-susceptible (n = 1893)
20-32
25
20-35
25
21-40
30
21-40
28-35*
Ampicillin-resistant beta-laetamase positive (n = 191)
10-30
25
10-34
24-28*
16--40
28
21-40
30-32*
Ampieitlin-resistant beta-laetamase negative
6-28
15-20"
6-28
15-20"
20-37
25
18--35
27-30*
(n = 128) *Broad modal peak.
The lack of isolates resistant to either cefixime or cefetamet made it impossible to perform valid regression analyses of zone and MIC data in order to suggest appropriate breakpoints. However, the distribution of isolates indicated that Haemophilus influenzae isolates inhibited only by > 1 mg/l cefetamet and >__ 0.5 mg/1 cefixime were markedly at variance with the susceptible population. Inhibition zone diameters for these two groups of isolates were generally < 22 mm for both drugs. For cefaclor and loracarbef, zone diameters < 20 mm correlated well with MICs _>16 mg/1. However, some isolates resistant by MIC parameters produced zone diameters > 20 mm and _< 25 mm, indicating relatively poor discriminatory values of these two discs. The prevalence of 13-1actamase production among Haemophilus influenzae in 1991 (21% type b and 8.3 % non-capsulated) compares well with 18 % and 6 % respectively in 1986 (2) and with similar average differences observed in Europe in 1986 (15 % and 10 %) and the USA in 1989 (29.5 % and 15 %) (6, 7). Non-13-1actamase-mediated resistance to ampicillin (MIC 1-64 mg/1) also increased from 4 % in 1986 to 5.8 % in 1991. While treatment with non-13-1actam drugs is an alternative, the popularity and the record of safety and efficacy of the 13-1actam agents has led to the development of several new drugs in this class in an attempt to combat these problems of resistance. Results of susceptibility testing with other 13-1actam drugs showed the expected shift upwards in MIC50 and MIC90 parameters of all four agents for the ampicillin-resistant, 15-1actamase-negative
group. While this phenomenon did not result in resistance to either cefetamet or cefixime according to currently prescribed breakpoints, organisms in this group accounted for 23 of the 25 isolates inhibited by > 1 mg cefetamet and for all eight isolates inhibited by > 0.5 mg/1 cefixime. Similarly, 80 (53 %) of the 150 isolates resistant to cefaclor and 57 (67 %) of the 85 isolates resistant to loracarbef were of this group. These findings agree with 1986 observations for cefaclor and cefixime (8). While the activity of loracarbef and cefetamet has been assessed for ampicillin-susceptible and 13-1actamase-positive isolates (9, 10), information on susceptibility has not previously been accumulated for a large collection of isolates with non-I~-lactamase-mediated resistance to ampicillin. The 6,8 % of isolates resistant to cefaclor, compared with 2 % in 1986 and 5.2 % among sputum isolates in 1990 (1), and the 3.8 % of isolates resistant to loracarbef indicate a need for reliable in vitro susceptibility testing of these agents. The lack of either cefetamet or cefixime-resistant isolates is encouraging, although this fact meant that the discriminative capacity of the 10 ~tg cefetamet disc could not be fully assessed. A 5 lag cefixime disc was adopted following earlier observations that the 30/ag disc initially produced was a poor indicator of susceptibility of Haemophilus influenzae and many other species (11). While results of this survey indicated that there are fewer problems of antimicrobial resistance in England and Scotland compared with several other European countries and parts of the USA, there is a greater possibility of encountering resis-
VO1. 11, 1992
tance to ampicillin and cefaclor than one decade ago. In particular, the increasing prevalence of non-15-1actamase-mediated resistance to ampicillin commonly results in resistance to some l~-lactam agents and a marked reduction in susceptibility to several others. However, insufficient information is available on the in vivo response of isolates to agents when MICs are near to but not above the recommended breakpoints to comment further on the clinical importance of this in vitro phenomenon.
References 1. Powell M, McVey D, Kassim MH, Chen HY, Williams JD: Antimierobial susceptibility of Streptococcuspneumoniae, Haemophilus influenzae and Moraxella (Branhamella) catarrhalis isolated, in the UK from sputa. Journal of Antimicrobial Chemotherapy 1991, 28: 249259.
2. Powell M, Koutsia-Carouzou C, Vouts'masD, Seymour A, Williams JD: Resistance of clinical isolates of Haemophilus influenzae in United Kingdom 1986. British Medical Journal 1987, 295: 176-179. 3. Howard A J, Williams HM: The prevalence of antibiotic resistance in Haemophilus influenzae in Wales. Journal of Antimicrobial Chemotherapy 21,1988: 251260. 4. Howard A J, Williams HM: The prevalence of antibiotic resistance in Haemophilus influenzae in Ireland. Journal of Antimierobial Chemotherapy 1989, 24: 963971. 5. Catlin BW: Iodometric detection of Haernophilus influenzae beta-lactamase: rapid presumptive test for ampicillin resistance. Antimierobial Agents and Chemotherapy 1975, 7: 265-270.
6. Machka K, Bmveny I, Dabernat H, Dornbusch K, Van Dyck E, Kayser FH, Van Klingeren B, Mittermayer H, Perea E, Powell M: Distribution and resistance patterns of Haemophilus influenzae: a European cooperative study. European Journal of Clinical Microbiology and Infectious Diseases 1988, 7: 14-24. 7. Jorgensen JH, Doern GC, Maher LA, Howell AW, Redding JS: Antimicrobial resistance among respiratory isolates of Haemophilus influenzae, Moraxella catarrhalis and Streptococcus pneumoniae in the United States. Antimicrobial Agents and Chemotherapy 1990, 34: 2075-2080. 8. Pnwell M, Williams JD: In vitro activity of cefaclor, cephalexin and ampicillin against 2,458 clinical isolates of Haemophilus influenzae. Journal of Antimicrobial Chemotherapy 21: 25-31. 9. Wise R, Andrews JM, Piddock LJV: In vitro activity of RO 15-8074 and RO 19-5247, two orally-administered cephalosporin metabolites. Antimicrobial Agents and Chemotherapy 1986, 29: 1067-1072.
10. Doern GV, Vautour R, Parker D, Tubert T, Tortes B: In vitro activity of loracarbef (LY 163891), a new oral earbacephem antimicrobial agent, against respiratory isolates of Haemophilus influenzae and Moraxella catarrhalis. Antimicrobial Agents and Chemotherapy 1991, 35: 1504-1507.
751
11. Fuchs PC, Barry AL, Jones RaN: Cefixime disk susceptibility test criteria. Journal of Clinical Microbiology 1986, 24: 647-649.
iiii
Antimicrobial Activity of Ten Macrolide, Linsosamine and Streptogramin Drugs Tested Against Legionella Species D.M. Johnson, M.E. Erwin, M.S. Barrett, B. Briggs Gooding, R.N. Jones*
RP59500, a semisynthetie pristinamycin combination, and 14 other antimicrobiai agents were tested against 108 Legionella strains. Of the ten macrolide, lincosamine and streptogramin agents tested, the new streptogramin RP59500 ranked seventh in order of activity against Legionella pneumophlla on the basis of MIC90 results as follows: clarithromycin - 14-OH clarithromycin ( M I C 9 0 0.12 rag/i) > roxithromycin > erythromycin = tylosin = virginiamycin > RP59500 (MIC90 1 mg/i) = azithromycin > dirithromycin > clindamycin (MIC90 8 mg/l). Of all 14 drugs tested in this study, rifampicin was the most potent with an MIC90 of 0.008 mg/i. In this retrospective study of Legionella strains (19811990 isolates), we observed no trend toward resistance to the agents investigated.
Since the first recognized epidemic in 1976, Legionelta pneumophila has been recognized as an important cause of nosocomial pneumonia (1). Legionella species are generally ubiquitous in the environment and have the ability to replicate within human phagocytic cells (2). Therefore, effective antibiotics must be able to penetrate these cells at biologically active concentrations. Rifampicin and erythromycin, the drugs of first choice, have demonstrated excellent cell penetration (3). In addition, other macrolides and tetracyclines appear to be promising therapeutic alternatives (4, 5, P. Rajagopalan-Levasseur et al., International Conference on the Macrolides, Azalides Department of Pathology, 5232 RCP, University of Iowa College of Medicine, Iowa City, Iowa 52242 USA.
New Antimicrobial Agents
In Vitro Susceptibility of
Haemophilus influenzaeto Cefaclor, Cefixime, Cefetamet and Loracarbef J.D. Williams*, M. Powell, Y.S. Fah, A. Seymour, M. Yuan
The susceptibility of 2,212 Haemophilus influenzae isolates cultured in UK clinical laboratories in 1991 was determined for four orally-administered [3-1actam drugs. These isolates included 1,893 ampicillin-susceptible, 191 I~-lactamase-positive and 128 ampiciilin-resistant, [3-1actamase-negative Haemophilus influenzae. While 150 (6.8 %) isolates were resistant to cefaclor (MIC > 16 rag/I) and 85 (3.8 %) to loracarbef, all were inhibited by < 2 mg/! cefetamet and _ 0.5 mg/l cefixime showed this type of resistance to ampicillin. Results indicate the importance of detecting non[3-1actamase-mediated resistance to ampicillin and any concomitant diminished susceptibility to other 13-1actam drugs.
In 1990 a UK survey of sputum isolates showed that the prevalence among 13-1actamase-positive and I~-lactamase-negative isolates of resistance to ampicillin (9.4 % and 5.2 % respectively) (1) had increased since the last survey of England and Scotland (6.2 % and 4 %) conducted in 1986 (2). The higher percentage of l~-lactamase-positive isolates noted was similar to figures obtained in Departmentof MedicalMicrobiology,The LondonHospital Medical College,Turner Street, London E1 2AD, UK.
Vol. 11. No. 8
Wales and Ireland of 8.4 % and 10.2 % respectively (3, 4). However, non-I~-lactamase-mediated resistance to ampicillin was not detected by UK laboratories in 85 of 100 such isolates collected during the 1986 survey. This is of importance because the mechanisms involved affect not only susceptibility to ampicillin but also commonly produce a marked degree of diminished susceptibility to most penicillins and cephalosporins with corresponding therapeutic implications. For example, the reduction in susceptibility to cefaclor and cefixime shown by such organisms was reported for the 100 13-1actamase-negative isolates obtained in the 1986 survey, which required 1-64 mg/l ampicillin for inhibition. While only 2 % of the total number of organisms collected (2458) were resistant to cefaclor, 47 of 50 cephalexin-resistant isolates were ampicillinresistant and ~-lactamase negative, as were all four isolates which were resistant to ceflxime. The modal MICs of the two cephalosporins for organisms with this mechanism of resistance were two to three doubling dilutions above modal values for the ampicillin-susceptible and 13-1actamase-positive groups. In the present study the 2,212 clinical isolates of
Haemophilus influenzae were examined for their susceptibility to three orally administered cephalosporins (cefaclor, cefixime and cefetamet) and to the carbacephem loracarbef. Susceptibility to these 15-1actam drugs was compared between groups of Haemophilus influenzae delineated according to their susceptibility to ampicillin.
Materials and Methods. All media, supplements and discs were obtained from Unipath, UK. Ampicillin was obtained from Sigma Chemical, USA, amoxicillin-clavulanate (in the form of'Adatabs') from Mast Laboratories, UK, cefaclor and loracarbef from Lilly Industries, UK, cefetamet from Roche Products, U K and cefixime from Ledefle Laboratories, UK. Methods were those used in previous studies (2). Briefly, five-hour cultures of organisms in brain heart infusion broth supplemented with 5 % (vol/vol) Fildes' supplement were diluted (1/100) in peptone water. The suspension was used for swab inoculation of several DST agar plates, supplemented with 0.25 % (vol/vol) lysed horse blood and 10 mg/1 NAD, to which the following discs were applied: cefaclor 30 lag, loracarbef 30 gg, cefetamet 10 gg and cefixime 5 ~g. The same suspension provided an inoculum of 104 cfu/ spot when deposited by a multipoint inoculator onto similar agar containing antimicrobial agents
Vol. 11, 1992
749
at final concentrations ranging from 256 to 0.008 mg/l. Incubation was for 18 h at 37 °C in an atmosp h e r e of 5 % CO2 and 95 % air. Control strains Haemophilus influenzae A T C C 491002 (fully-susceptible), Haemophilus influenzae W10311 ([3lactamase-positive, chloramphenicol and tetracycline-resistant) and Haemophilus influenzae 518 (ampicillin-resistant, 13-1actamase-negative) were included in all batches tested. Production of I~-lactamase was detected by the iodometric m e t h o d (5).
Results and Discussion. Using the same inhibition zone diameter and MIC breakpoints adopted in previous studies and those suggested by manufacturers in the case of cefetamet and loracarbef, the numbers o f isolates resistant to the agents tested are shown in Table 1. The susceptibility of all isolates to the three cephalosporins and to loracarbef is c o m p a r e d in Table 2, in which MIC50 and MIC90 values are given for organisms grouped according to their susceptibility to am-
Table 1:
picillin. As expected, a rise in b o t h values of two to three doubling dilutions was seen for three drugs between the ampicillin-susceptible group and those organisms which showed non-l~-lactamase-mediated diminished susceptibility to ampicillin. T h e exception was cefixime although the reason for this was partly the fact that the distribution of all isolates according to the M I C of cefixime was generally in a broader, flatter curve compared with other 15-1actam drugs. Table 3 shows similar grouping of isolates according to inhibition zone diameter, comparing range and modal values for the four l~-lactam antimicrobial agents. While there was little discernible difference in modal values between the ampicillin-susceptible and 13-1actamase-positive groups, the range extended downwards for the latter group except in the case o f cefixime. For all four agents these two parameters shifted downwards in the I~-lactamase-negative, ampicillinresistant group.
Preva~ence~fresistancet~ampici~inandf~ur~therr3-~actamagentsam~ng2~2~2Haem~philusin~uenzaeis~ates.
Antimicrobial agent
Criteria for resistance MIC (mg/l)
Ampieillin Ampicillin Cefaelor Loracarbef Cefetamet Cefixime
Number (%) resistant 1991
not applicable >1 _>16 > 16 >8 >2
191a 128b 150 85 none none
(8.6) (5.8) (6.8) (3.8) (0) (0)
aBeta-lactamase positive isolates. bBeta-laetamase negative isolates.
Table 2: Comparison of the susceptibility of 2,212 Haemophilusinfluenzaeisolates to ampicillin and to four other I%lactam antimicrobial agents. MIC (mgll) Susceptibility to ampieillin
Cefaclor
Loracarbef
Cefetamet
Cefixime
MIC50
MIC90
MIC50
MIC90
MIC50
MIC90
MIC50
MIC90
Ampicillin-susceptible (n=1893)
4
8
2
4
0.12
0.25
0.06
0.12
Ampicillin-resistant beta-lactamase positive
4
8
2
8
0.25
0.5
0.03
0.06
16
64
8
32
0.5
1
0.06
0.25
(n =
191)
Ampicillin-resistant beta-taetamase negative (n = 128)
750
Eur. J. Clin. Microbiol. Infect. Dis.
Table 3: Comparison of the susceptibility of 2,212 Haemophilus influenzae isolates to ampiciUin and to four other 13-1attain
antimierobial agents according to inhibition zone diameters. III II II IIIIIII IIIII II
hibition zone diameters (mm) Susceptibility to ampieillin
Cefaclor
Loraearbef
Cefetamet
Cefixime
Range
Mode
Range
Mode
Range
Mode
Range
Mode
Ampi¢illin-susceptible (n = 1893)
20-32
25
20-35
25
21-40
30
21-40
28-35*
Ampicillin-resistant beta-laetamase positive (n = 191)
10-30
25
10-34
24-28*
16--40
28
21-40
30-32*
Ampieitlin-resistant beta-laetamase negative
6-28
15-20"
6-28
15-20"
20-37
25
18--35
27-30*
(n = 128) *Broad modal peak.
The lack of isolates resistant to either cefixime or cefetamet made it impossible to perform valid regression analyses of zone and MIC data in order to suggest appropriate breakpoints. However, the distribution of isolates indicated that Haemophilus influenzae isolates inhibited only by > 1 mg/l cefetamet and >__ 0.5 mg/1 cefixime were markedly at variance with the susceptible population. Inhibition zone diameters for these two groups of isolates were generally < 22 mm for both drugs. For cefaclor and loracarbef, zone diameters < 20 mm correlated well with MICs _>16 mg/1. However, some isolates resistant by MIC parameters produced zone diameters > 20 mm and _< 25 mm, indicating relatively poor discriminatory values of these two discs. The prevalence of 13-1actamase production among Haemophilus influenzae in 1991 (21% type b and 8.3 % non-capsulated) compares well with 18 % and 6 % respectively in 1986 (2) and with similar average differences observed in Europe in 1986 (15 % and 10 %) and the USA in 1989 (29.5 % and 15 %) (6, 7). Non-13-1actamase-mediated resistance to ampicillin (MIC 1-64 mg/1) also increased from 4 % in 1986 to 5.8 % in 1991. While treatment with non-13-1actam drugs is an alternative, the popularity and the record of safety and efficacy of the 13-1actam agents has led to the development of several new drugs in this class in an attempt to combat these problems of resistance. Results of susceptibility testing with other 13-1actam drugs showed the expected shift upwards in MIC50 and MIC90 parameters of all four agents for the ampicillin-resistant, 15-1actamase-negative
group. While this phenomenon did not result in resistance to either cefetamet or cefixime according to currently prescribed breakpoints, organisms in this group accounted for 23 of the 25 isolates inhibited by > 1 mg cefetamet and for all eight isolates inhibited by > 0.5 mg/1 cefixime. Similarly, 80 (53 %) of the 150 isolates resistant to cefaclor and 57 (67 %) of the 85 isolates resistant to loracarbef were of this group. These findings agree with 1986 observations for cefaclor and cefixime (8). While the activity of loracarbef and cefetamet has been assessed for ampicillin-susceptible and 13-1actamase-positive isolates (9, 10), information on susceptibility has not previously been accumulated for a large collection of isolates with non-I~-lactamase-mediated resistance to ampicillin. The 6,8 % of isolates resistant to cefaclor, compared with 2 % in 1986 and 5.2 % among sputum isolates in 1990 (1), and the 3.8 % of isolates resistant to loracarbef indicate a need for reliable in vitro susceptibility testing of these agents. The lack of either cefetamet or cefixime-resistant isolates is encouraging, although this fact meant that the discriminative capacity of the 10 ~tg cefetamet disc could not be fully assessed. A 5 lag cefixime disc was adopted following earlier observations that the 30/ag disc initially produced was a poor indicator of susceptibility of Haemophilus influenzae and many other species (11). While results of this survey indicated that there are fewer problems of antimicrobial resistance in England and Scotland compared with several other European countries and parts of the USA, there is a greater possibility of encountering resis-
VO1. 11, 1992
tance to ampicillin and cefaclor than one decade ago. In particular, the increasing prevalence of non-15-1actamase-mediated resistance to ampicillin commonly results in resistance to some l~-lactam agents and a marked reduction in susceptibility to several others. However, insufficient information is available on the in vivo response of isolates to agents when MICs are near to but not above the recommended breakpoints to comment further on the clinical importance of this in vitro phenomenon.
References 1. Powell M, McVey D, Kassim MH, Chen HY, Williams JD: Antimierobial susceptibility of Streptococcuspneumoniae, Haemophilus influenzae and Moraxella (Branhamella) catarrhalis isolated, in the UK from sputa. Journal of Antimicrobial Chemotherapy 1991, 28: 249259.
2. Powell M, Koutsia-Carouzou C, Vouts'masD, Seymour A, Williams JD: Resistance of clinical isolates of Haemophilus influenzae in United Kingdom 1986. British Medical Journal 1987, 295: 176-179. 3. Howard A J, Williams HM: The prevalence of antibiotic resistance in Haemophilus influenzae in Wales. Journal of Antimicrobial Chemotherapy 21,1988: 251260. 4. Howard A J, Williams HM: The prevalence of antibiotic resistance in Haemophilus influenzae in Ireland. Journal of Antimierobial Chemotherapy 1989, 24: 963971. 5. Catlin BW: Iodometric detection of Haernophilus influenzae beta-lactamase: rapid presumptive test for ampicillin resistance. Antimierobial Agents and Chemotherapy 1975, 7: 265-270.
6. Machka K, Bmveny I, Dabernat H, Dornbusch K, Van Dyck E, Kayser FH, Van Klingeren B, Mittermayer H, Perea E, Powell M: Distribution and resistance patterns of Haemophilus influenzae: a European cooperative study. European Journal of Clinical Microbiology and Infectious Diseases 1988, 7: 14-24. 7. Jorgensen JH, Doern GC, Maher LA, Howell AW, Redding JS: Antimicrobial resistance among respiratory isolates of Haemophilus influenzae, Moraxella catarrhalis and Streptococcus pneumoniae in the United States. Antimicrobial Agents and Chemotherapy 1990, 34: 2075-2080. 8. Pnwell M, Williams JD: In vitro activity of cefaclor, cephalexin and ampicillin against 2,458 clinical isolates of Haemophilus influenzae. Journal of Antimicrobial Chemotherapy 21: 25-31. 9. Wise R, Andrews JM, Piddock LJV: In vitro activity of RO 15-8074 and RO 19-5247, two orally-administered cephalosporin metabolites. Antimicrobial Agents and Chemotherapy 1986, 29: 1067-1072.
10. Doern GV, Vautour R, Parker D, Tubert T, Tortes B: In vitro activity of loracarbef (LY 163891), a new oral earbacephem antimicrobial agent, against respiratory isolates of Haemophilus influenzae and Moraxella catarrhalis. Antimicrobial Agents and Chemotherapy 1991, 35: 1504-1507.
751
11. Fuchs PC, Barry AL, Jones RaN: Cefixime disk susceptibility test criteria. Journal of Clinical Microbiology 1986, 24: 647-649.
iiii
Antimicrobial Activity of Ten Macrolide, Linsosamine and Streptogramin Drugs Tested Against Legionella Species D.M. Johnson, M.E. Erwin, M.S. Barrett, B. Briggs Gooding, R.N. Jones*
RP59500, a semisynthetie pristinamycin combination, and 14 other antimicrobiai agents were tested against 108 Legionella strains. Of the ten macrolide, lincosamine and streptogramin agents tested, the new streptogramin RP59500 ranked seventh in order of activity against Legionella pneumophlla on the basis of MIC90 results as follows: clarithromycin - 14-OH clarithromycin ( M I C 9 0 0.12 rag/i) > roxithromycin > erythromycin = tylosin = virginiamycin > RP59500 (MIC90 1 mg/i) = azithromycin > dirithromycin > clindamycin (MIC90 8 mg/l). Of all 14 drugs tested in this study, rifampicin was the most potent with an MIC90 of 0.008 mg/i. In this retrospective study of Legionella strains (19811990 isolates), we observed no trend toward resistance to the agents investigated.
Since the first recognized epidemic in 1976, Legionelta pneumophila has been recognized as an important cause of nosocomial pneumonia (1). Legionella species are generally ubiquitous in the environment and have the ability to replicate within human phagocytic cells (2). Therefore, effective antibiotics must be able to penetrate these cells at biologically active concentrations. Rifampicin and erythromycin, the drugs of first choice, have demonstrated excellent cell penetration (3). In addition, other macrolides and tetracyclines appear to be promising therapeutic alternatives (4, 5, P. Rajagopalan-Levasseur et al., International Conference on the Macrolides, Azalides Department of Pathology, 5232 RCP, University of Iowa College of Medicine, Iowa City, Iowa 52242 USA.
