and Health Sciences, Wayne State University and Detroit Receiving Hospital, Detroit, MI, USA

Journal of Antimicrobial Chemotherapy (2004) 54, Suppl. S1, i7 i15 DOI: 10.1093/jac/dkh313 JAC Antimicrobial susceptibility of Streptococcus pneumoniae, Streptococcus pyogenes and Haemophilus influenzae collected from patients across the USA, in 2001 2002, as part of the PROTEKT US study Steven D. Brown 1 * and Michael J. Rybak 2 1 Clinical Microbiology Institute, 9725 SW Commerce Circle, Suite A1, Wilsonville, OR 97070; 2 Anti-Infective Research Laboratory, Department of Pharmacy Practice, Eugene Applebaum College of Pharmacy and Health Sciences, Wayne State University and Detroit Receiving Hospital, Detroit, MI, USA Introduction Background: The PROTEKT US (Prospective Resistant Organism Tracking and Epidemiology for the Ketolide Telithromycin in the United States) surveillance programme was started in 2000, to chart the emergence and spread of antimicrobial resistance among isolates of Streptococcus pneumoniae, Streptococcus pyogenes and Haemophilus influenzae from across the USA. Methods: In 2001 2002 (Year 2 of PROTEKT US) 242 centres from 46 states and the territory of Puerto Rico submitted a total of 10 012 S. pneumoniae, 4508 S. pyogenes and 3296 H. influenzae isolates from community-acquired respiratory tract infections (CARTIs). Susceptibility testing was performed and interpreted using NCCLS methodology and criteria. Results: Overall, 35.4% of S. pneumoniae were non-susceptible to penicillin (14.2% intermediate, MIC 0.12 1 mg/l; 21.2% resistant, MIC >_2 mg/l) and 27.9% were resistant to erythromycin (MIC >_1 mg/l) (0.2% intermediate, MIC 0.5 mg/l). A total of 105 (1.0%) isolates were resistant to levofloxacin (MIC >_ 8 mg/l). More than 99.2% of isolates were susceptible to telithromycin (MIC <_ 1 mg/l) irrespective of penicillin and/or erythromycin resistance. All S. pyogenes isolates were susceptible to penicillin (MIC <_ 0.12 mg/l) and 5.7% were resistant to erythromycin (MIC >_ 1 mg/l) (0.3% intermediate, MIC 0.5 mg/l). The MIC 90 of telithromycin for S. pyogenes was 0.03 mg/l. A total of 27.5% of H. influenzae isolates were b-lactamase producers. Overall, 27.8% were resistant (MIC >_ 4 mg/l) and 1.1% were intermediate to ampicillin (MIC 2 mg/l). A total of 96.3% of H. influenzae isolates were susceptible to telithromycin (MIC <_4 mg/l). Conclusions: Antimicrobial resistance continues to be a problem in the USA. The ketolide telithromycin continues to show high activity against common CARTI pathogens, including those resistant to b-lactams and macrolides. Keywords: telithromycin, CARTIs, S. pneumoniae Antimicrobial resistance among community-acquired respiratory tract infection (CARTI) pathogens is a continuing problem in the USA as it is globally. 1 In this climate of antimicrobial resistance, surveillance studies have become increasingly important, with the focus on obtaining not only national but also regional and local data. The ongoing PROTEKT US (Prospective Resistant Organism Tracking and Epidemiology for the Ketolide Telithromycin in the United States) surveillance study, which began in 2000, is designed to meet these needs, through the longitudinal monitoring of antimicrobial resistance among Streptococcus pneumoniae, Streptococcus pyogenes and Haemophilus influenzae isolates collected from CARTIs. Here, we report findings from the second year (2001 2002) of PRO- TEKT US, charting susceptibility among isolates of these three bacterial species collected from the 242 centres across the USA. Materials and methods Bacterial isolate collection In the second year (2001 2002) of PROTEKT US, bacterial isolates were submitted from 242 centres from 46 states and Puerto Rico. These states were distributed across the six regions of PROTEKT US (Figure 1). Each centre was requested to submit at least... *Corresponding author. Tel: +1-503-682-3232; Fax: +1-503-682-4548; E-mail: SBrown@clinmicroinst.com... i7 JAC q The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.

S. D. Brown and M. J. Rybak Figure 1. States involved in the second year (2001 2002) of PROTEKT US. 60 S. pneumoniae isolates, 20 S. pyogenes and 15 H. influenzae, collected from patients with community-acquired pneumonia, acute bacterial exacerbation of chronic bronchitis, acute exacerbation of chronic obstructive pulmonary disease, acute/chronic sinusitis, acute/chronic otitis media or acute/chronic pharyngitis. Acceptable culture sources were blood, sputum, bronchoalveolar lavage, middle ear fluid, nasopharyngeal swab or aspirate cultures, sinus aspirate cultures or, for S. pyogenes only, throat cultures. Antimicrobial susceptibility testing Isolates were shipped to a central laboratory (CMI, Wilsonville, OR, USA) and, after re-identification, MICs were determined according to the NCCLS broth microdilution methodology. 2 Isolates of S. pneumoniae and S. pyogenes were tested against penicillin, co-amoxiclav, cefuroxime, clindamycin, erythromycin, clarithromycin, azithromycin, telithromycin, levofloxacin, gatifloxacin, tetracycline, linezolid and co-trimoxazole. H. influenzae isolates were tested against ampicillin, co-amoxiclav, cefprozil, cefuroxime, cefotaxime, clarithromycin, azithromycin, telithromycin, levofloxacin, gatifloxacin, moxifloxacin, tetracycline and linezolid (data not shown). b-lactamase production was determined for isolates of H. influenzae using the chromogenic cephalosporin (nitrocefin) test (Oxoid, Basingstoke, UK). Susceptibility was assessed according to NCCLS breakpoints. 3 For telithromycin the tentative NCCLS 2004 breakpoints 4 were applied to S. pneumoniae: susceptible <_ 1 mg/l, intermediate 2 mg/l and resistant >_4 mg/l; and H. influenzae: susceptible <_4 mg/l, intermediate 8 mg/l and resistant >_ 16 mg/l. No telithromycin breakpoints are currently proposed for S. pyogenes. Results Streptococcus pneumoniae Penicillin susceptibility. During 2001 2002, 10 012 isolates of S. pneumoniae were collected from across the USA. Overall, 35.4% of isolates were non-susceptible to penicillin (MIC >_ 0.12 mg/l) with 21.2% fully resistant (MIC >_2 mg/l). The highest rate of penicillin non-susceptibility was detected in the south-central region (44.3%, Table 1), where 28.9% of isolates were fully resistant to penicillin. A total of 65 isolates were collected with a penicillin MIC of 8 mg/l and one isolate with a penicillin MIC of 16 mg/l. These were collected from 26 states, across all six regions. Variation was noted in the prevalence of penicillin non-susceptibility according to the culture source of the isolates (Table 1). Almost all penicillin-resistant isolates of S. pneumoniae were resistant to cefuroxime (0.1% susceptible) and one-third showed reduced susceptibility to co-amoxiclav (67.1% susceptible, Table 2). Susceptibility to the macrolides was reduced among penicillin-resistant isolates, with only 19.4% of isolates remaining susceptible to erythromycin. Clindamycin, tetracycline and co-trimoxazole also had reduced activity against penicillinresistant S. pneumoniae to varying extents (74.7%, 52.6% and 8.1% susceptibility, respectively). Susceptibility of penicillinintermediate S. pneumoniae was reduced to cefuroxime (67.0%), macrolides (erythromycin 50.1%), tetracycline (77.2%) and co-trimoxazole (48.0%). The most active antimicrobials against penicillin-resistant S. pneumoniae, according to MIC 90, were telithromycin and gatifloxacin (MIC 90 0.5 mg/l, Table 2). Susceptibility among penicillin-resistant isolates was highest to gatifloxacin, telithromycin and linezolid (98.1%, 99.4% and 99.9%, respectively). Telithromycin was the most active antimicrobial against penicillin-intermediate isolates (MIC 90 0.25 mg/l). Susceptibility among penicillin-intermediate isolates was highest to telithromycin, co-amoxiclav and linezolid (99.9%, 99.9% and 100%, respectively). The MIC distributions of penicillin, erythromycin, telithromycin and levofloxacin for penicillin-intermediate and -resistant S. pneumoniae are shown in Figure 2. Macrolide susceptibility. A total of 27.9% of isolates were resistant to erythromycin (representative macrolide). Distribution of i8

i9 Table 1. Distribution of penicillin, macrolide (erythromycin) and fluoroquinolone (levofloxacin) resistance among S. pneumoniae isolates from across the USA according to geographical region and culture source n susceptible (MIC <_1 mg/l) Percentage of S. pneumoniae isolates penicillin macrolide a fluoroquinolone b intermediate (MIC 0.12 1 mg/l) resistant (MIC >_2 mg/l) susceptible (MIC <_0.25 mg/l resistant (MIC >_1 mg/l) susceptible (MIC <_2 mg/l) resistant (MIC >_8 mg/l) Region total USA 10 012 64.6 14.2 21.2 71.9 27.9 98.8 1.0 north-central 2177 64.3 14.2 21.5 71.0 28.8 99.0 0.9 north-east 3168 69.6 13.2 17.2 75.3 24.3 98.7 1.3 north-west 531 72.7 11.7 15.6 77.2 22.4 99.6 0.4 south-central 1333 55.7 15.4 28.9 65.3 34.7 98.7 1.2 south-east 1392 57.2 16.2 26.6 65.4 34.6 99.3 0.6 south-west 1411 66.3 14.5 19.2 76.3 23.6 98.4 1.3 Culture source total 9531 64.5 14.0 21.5 72.0 27.8 98.8 1.1 sputum 4115 66.0 14.7 19.3 74.3 25.5 98.1 1.7 blood 2717 73.1 11.0 15.9 78.9 20.9 99.5 0.5 BAL c 1019 56.4 16.7 26.9 65.6 34.2 98.9 1.1 ear 689 50.6 16.7 32.7 59.8 39.9 99.7 0.1 NAP c 601 51.6 15.1 33.3 58.4 41.6 99.7 0.3 sinus 390 53.8 13.6 32.6 59.2 40.8 98.5 1.5 a A total of 18 isolates were macrolide intermediate (MIC 0.5 mg/l), data not shown, therefore percentages may not add up to 100. b A total of 11 isolates were levofloxacin intermediate (MIC 4 mg/l), data not shown, therefore percentages may not add up to 100. c BAL, bronchoalveolar lavage; NAP, nasopharynx. Susceptibility of RTI pathogens in 2001 2002 in the USA

S. D. Brown and M. J. Rybak Table 2. Antimicrobial activity and susceptibility among S. pneumoniae isolates from across the USA, according to penicillin and macrolide susceptibility All isolates (n = 10 012) PEN-I a (n = 1424) PEN-R a (n = 2124) MAC-R a (n = 2793) MIC90 MIC90 MIC90 MIC90 Antimicrobial Penicillin 2 <_ 0.06 16 64.6 1 0.12 1 0.0 4 2 16 0.0 4 <_ 0.06 16 13.5 Co-amoxiclav 2 <_ 0.12 >4 93.0 1 <_ 0.12 4 99.9 >4 0.25 >4 67.1 >4 <_ 0.12 >4 77.8 Cefuroxime 8 <_ 0.12 >8 74.1 4 <_ 0.12 >8 67.0 >8 1 >8 0.1 8 <_ 0.12 >8 29.1 Clindamycin <_ 0.25 <_ 0.25 >64 91.5 >64 <_ 0.25 >64 86.9 >64 <_ 0.25 >64 74.7 >64 <_ 0.25 >64 70.0 Erythromycin 16 <_ 0.06 >256 71.9 >256 <_ 0.06 >256 50.1 >256 <_ 0.06 >256 19.4 >256 1 >256 0.0 Clarithromycin 16 <_ 0.03 >256 72.1 >256 <_ 0.03 >256 50.2 >256 <_ 0.03 >256 19.7 >256 0.12 >256 0.3 Azithromycin 32 <_ 0.03 >256 71.9 >256 <_ 0.03 >256 50.2 >256 <_ 0.03 >256 19.6 >256 0.5 >256 0.1 Telithromycin 0.25 <_ 0.015 4 99.9 0.25 <_ 0.015 2 99.9 0.5 <_ 0.015 4 99.4 0.5 <_ 0.015 4 99.5 Levofloxacin 1 <_ 0.12 256 98.8 1 <_ 0.12 128 97.8 1 0.5 256 97.8 1 0.25 256 97.4 Gatifloxacin 0.5 <_ 0.06 >4 98.9 0.5 <_ 0.06 >4 97.9 0.5 <_ 0.06 >4 98.1 0.5 <_ 0.06 >4 97.6 Tetracycline >4 <_ 0.06 >4 84.9 >4 <_ 0.06 >4 77.2 >4 <_ 0.06 >4 52.6 >4 <_ 0.06 >4 51.9 Linezolid 2 <_ 0.06 4 99.9 2 0.12 2 100 2 0.5 4 99.9 2 0.12 4 99.9 Co-trimoxazole >4 <_ 0.25 >4 65.2 >4 <_ 0.25 >4 48.0 >4 <_ 0.25 >4 8.1 >4 <_ 0.25 >4 18.5 a PEN-I, penicillin intermediate, MIC 0.12 1 mg/l; PEN-R, penicillin resistant, MIC >_2 mg/l; MAC-R, erythromycin resistant, MIC >_1 mg/l. resistance by region followed a similar pattern to penicillin, with the highest rates of macrolide resistance detected in the southcentral and south-east regions (34.7% and 34.6%, respectively, Table 1). As with penicillin, variation in the prevalence of macrolide resistance was noted among the culture sources (Table 1). There was almost complete cross-resistance between the three macrolides tested, with only 0.1% and 0.3% of erythromycinresistant isolates remaining susceptible to azithromycin and clarithromycin, respectively; 70.0% of macrolide-resistant isolates were susceptible to clindamycin. Macrolide-resistant isolates showed reduced susceptibility to penicillin (13.5% susceptible), cefuroxime (29.1%), tetracycline (51.9%) and co-trimoxazole (18.5%) (Table 2). The most active agents against macrolide-resistant S. pneumoniae were again telithromycin and gatifloxacin (MIC 90 0.5 mg/l, Table 2), with the greatest percentage of isolates susceptible to telithromycin and linezolid (99.5% and 99.9%, respectively). The MIC distributions of penicillin, erythromycin, telithromycin and levofloxacin for macrolide (erythromycin)-resistant S. pneumoniae are shown in Figure 2. Fluoroquinolone susceptibility. A total of 105 (1.0%) S. pneumoniae isolates were resistant to levofloxacin (MIC >_ 8 mg/l), and 11 (0.1%) were intermediate (MIC 4 mg/l). In addition 215 (2.1%) isolates were at the levofloxacin susceptibility breakpoint of 2 mg/l. Resistance to gatifloxacin (MIC >_ 4 mg/l) was detected in 85 isolates (all of which were also resistant to levofloxacin). A further 22 isolates (0.2%) were gatifloxacin intermediate (MIC 2 mg/l) and 25 (0.2%) isolates were at the susceptibility breakpoint of 1 mg/l. Of the 20 levofloxacinresistant isolates that were not resistant to gatifloxacin, 19 were intermediate (MIC 2 mg/l) and one isolate was susceptible (MIC 0.5 mg/l). The highest rate of fluoroquinolone (levofloxacin) resistance was detected in the north-east and south-west regions (both 1.3%, Table 1). Susceptibility among isolates resistant to both penicillin and macrolides A total of 1709 (17.1%) isolates of S. pneumoniae were resistant to both penicillin and macrolides. Few of these isolates were susceptible to cefuroxime (0.2%) or co-trimoxazole (2.5%) (Table 3). Telithromycin (99.3% susceptible) and linezolid (99.8%) retained greatest activity against these isolates. Streptococcus pyogenes A total of 4508 isolates of S. pyogenes were collected during 2001 2002 and, of the 4496 isolates with culture source information, 4159 (92.5%) were from throat cultures. All isolates were susceptible to penicillin (Table 4). Overall, 5.7% of isolates were resistant (MIC >_ 1 mg/l) and 0.3% were intermediate (MIC 0.5 mg/l) to erythromycin, with similar results for clarithromycin (5.5% resistant, 0.1% intermediate) and azithromycin (5.6% resistant, 0.3% intermediate). The MIC 90 of telithromycin was 0.03 mg/l, which was lower than those of azithromycin (0.25 mg/l) and erythromycin (0.12 mg/l). The MIC range for telithromycin against S. pyogenes was <_ 0.015 16 with an MIC of >_4 mg/l for 12 isolates. These 12 isolates were also highly resistant to the macrolides (MICs >_ 256 mg/l). i10

Susceptibility of RTI pathogens in 2001 2002 in the USA Figure 2. MIC distributions of penicillin, erythromycin, telithromycin and levofloxacin for S. pneumoniae isolates grouped according to penicillin (PEN-I, penicillin intermediate, MIC 0.12 1 mg/l; PEN-R, penicillin resistant, MIC >_ 2 mg/l) and macrolide (MAC-R, erythromycin resistant, MIC >_ 1 mg/l) susceptibility. Key: black bars, all isolates (n = 10 012); striped bars, penicillin-intermediate isolates (n = 1424); white bars, penicillin-resistant isolates (n = 2124); grey bars, macrolide-resistant isolates (n = 2793). Macrolide (erythromycin) resistance was similar in the six regions of the USA, with the highest rate of resistance observed in the south-central region (7.0%, Table 5). Haemophilus influenzae Among 3296 H. influenzae isolates collected in 2001 2002, 905 (27.5%) were b-lactamase producers. The highest prevalence of b-lactamase-producing isolates was among ear cultures (41.2%, Table 6). b-lactamase-producing isolates were most common in the south-central region (33.4%, Table 6). Overall, 27.8% of H. influenzae isolates were resistant to ampicillin. None of the 905 b-lactamase-producing isolates was susceptible to ampicillin (0.2% intermediate and 99.8% resistant). Thirteen isolates were ampicillin resistant (MIC >_ 4 mg/l), but b-lactamase-negative (BLNAR); six were from the north-central i11

S. D. Brown and M. J. Rybak Table 3. Antimicrobial susceptibility among 1709 S. pneumoniae isolates resistant to both penicillin and macrolides from across the USA Antimicrobial MIC 90 (mg/l) Range (mg/l) %S Penicillin 4 2 16 0 Erythromycin >256 1 >256 0 Co-amoxiclav >4 0.25 >4 63.8 Cefuroxime >8 1 >8 0.2 Clindamycin >64 <_ 0.25 >64 68.6 Clarithromycin >256 0.12 >256 0.2 Azithromycin >256 0.5 >256 0.2 Telithromycin 1 <_ 0.015 4 99.3 Levofloxacin 1 0.5 256 97.4 Gatifloxacin 0.5 <_ 0.06 >4 97.7 Tetracycline >4 <_ 0.06 >4 42.5 Linezolid 2 0.5 4 99.8 Co-trimoxazole >4 <_ 0.25 >4 2.5 Table 4. Antimicrobial activity and susceptibility among 4508 S. pyogenes isolates from across the USA Antimicrobial MIC 90 (mg/l) Range (mg/l) %S Penicillin <_ 0.06 <_ 0.06 0.12 100 Co-amoxiclav <_ 0.12 <_ 0.12 0.25 predicted from penicillin Cefuroxime <_ 0.12 <_ 0.12 0.5 predicted from penicillin Clindamycin <_ 0.25 <_ 0.25 >64 99.4 Erythromycin 0.12 <_ 0.06 >256 94.0 Clarithromycin 0.06 <_ 0.03 >256 94.4 Azithromycin 0.25 <_ 0.03 >256 94.1 Telithromycin a 0.03 <_0.015 16 Levofloxacin 1 0.25 8 99.9 Gatifloxacin 0.25 <_ 0.06 4 99.9 Tetracycline 0.25 <_ 0.06 >4 92.0 Linezolid 2 <_ 0.06 2 100 Co-trimoxazole a 0.5 <_0.25 >4 a No NCCLS breakpoints available. region, three from the north-east, two from the south-east and two from the south-west. All BLNAR isolates were susceptible to cefotaxime, clarithromycin, azithromycin, telithromycin, levofloxacin, gatifloxacin and moxifloxacin; four were resistant to co-amoxiclav (nine susceptible); three were resistant to cefuroxime (two susceptible, eight intermediate); three were resistant to tetracycline (10 susceptible); one isolate was susceptible to cefprozil and 12 were intermediate or resistant (cefprozil test range 0.12 16 mg/l). With the exception of ampicillin, the susceptibility of H. influenzae isolates ranged from 80.2% for clarithromycin to 100% for cefotaxime (Table 7). Among b-lactamase producers, cefprozil and clarithromycin had the lowest activity (59.9% and 75.8% of isolates susceptible, respectively). The MICs of all three fluoroquinolones tested were above the NCCLS susceptibility breakpoints for 10 H. influenzae isolates (levofloxacin MIC >2 mg/l, moxifloxacin and gatifloxacin MICs >1 mg/l). Seven of the isolates were collected from sputum, one from bronchoalveolar lavage, one from blood and one from nasopharyngeal swab or aspirate culture. The age range of the patients was 39 85 years. All of these isolates were from New York State; nine were b-lactamase negative and remained susceptible to all the other antimicrobials tested. Discussion These data from 2001 2002 report similar percentages of S. pneumoniae isolates non-susceptible to penicillin and resistant to macrolides to those reported by others in recent years, 5 8 with the south-central and south-east regions having the highest percentages. In all regions, penicillin-resistant isolates were more prevalent than penicillin-intermediate isolates. Penicillin susceptibility also varied with the culture source of the S. pneumoniae isolates, with the highest rates of non-susceptibility among isolates from nasopharyngeal, ear and sinus cultures. This is in accordance with the findings of Doern et al. 5 and Jacobs et al. 9 Irrespective of culture source, penicillin resistance was more prevalent than penicillin intermediate resistance. A total of 65 isolates with a penicillin MIC of 8 mg/l and one isolate with a penicillin MIC of 16 mg/l were collected from across the USA. Isolates with penicillin MICs of >_ 8 mg/l have been reported previously in Table 5. Distribution of macrolide (erythromycin) resistance among 4508 S. pyogenes isolates from across the USA according to geographical region Region n susceptible (MIC <_0.25 mg/l) Percentage of S. pyogenes isolates intermediate (MIC 0.5 mg/l) resistant (MIC >_1 mg/l) Total USA 4508 94.0 0.3 5.7 North-central 991 94.1 0.0 5.9 North-east 1444 93.7 0.4 5.9 North-west a 230 93.0 0.4 6.5 South-central 617 93.0 0.0 7.0 South-east 610 93.9 0.5 5.6 South-west a 616 95.6 0.6 3.7 a Percentages add up to 99.9% due to rounding. i12

Susceptibility of RTI pathogens in 2001 2002 in the USA Table 6. Distribution of H. influenzae isolates from across the USA according to geographical region and culture source n Percentage of H. influenzae isolates b-lactamase negative (n = 2391) b-lactamase positive (n = 905) Region total USA 3296 72.5 27.5 north-central 720 71.9 28.1 north-east 1074 72.0 28.0 north-west 173 74.6 25.4 south-central 401 66.6 33.4 south-east 494 74.1 25.9 south-west 434 77.9 22.1 Culture source total 2965 73.4 26.6 sputum 1885 76.3 23.7 blood 126 69.0 31.0 BAL 411 71.5 28.5 ear 177 58.8 41.2 NAP 245 67.3 32.7 sinus 121 73.6 26.4 BAL, bronchoalveolar lavage; NAP, nasopharynx. the USA, 6,7,10 and a recent study by Jones et al. 11 recorded two isolates with MICs of 16 and >32 mg/l, respectively. Among penicillin-intermediate and -resistant isolates, susceptibility to cefuroxime, clindamycin, the macrolides, tetracycline and co-trimoxazole was reduced. Susceptibility to co-amoxiclav was not reduced among penicillin-intermediate isolates. The only antimicrobials to maintain susceptibility and activity against penicillin-intermediate and -resistant isolates were telithromycin, the fluoroquinolones and linezolid. Susceptibility was reduced to penicillin, clindamycin, tetracycline and co-trimoxazole among macrolide-resistant S. pneumoniae. Only telithromycin, the fluoroquinolones and linezolid maintained susceptibility and activity. The mechanism of macrolide resistance is typically deduced using the clindamycin and Table 7. Antimicrobial activity and susceptibility among H. influenzae isolates from across the USA Antimicrobial macrolide MICs. Such deductions will not be discussed here as the genotypes of the macrolide-resistant isolates are presented elsewhere in this supplement. 12 There has been much discussion in the literature regarding the emergence of fluoroquinolone resistance in S. pneumoniae. In 2001 2002, the PROTEKT US programme collected 105 (1.0%) isolates that were resistant to levofloxacin. These isolates were distributed across the six regions of PROTEKT US. This is similar to the rate reported by Karlowsky et al. 6 for the same season, and Doern & Brown 13 for the first year of PROTEKT US. Brueggemann et al. 14 reported a rate of 0.5% in the USA between 1994 and 2000. Fluoroquinolone resistance is typically mediated by mutations in the quinolone-resistance-determining regions of both gyra (DNA gyrase subunit) and parc (topoisomerase IV subunit). 15 A single mutation may reduce the binding affinity of fluoroquinolones, thereby reducing activity and increasing the MIC, but is insufficient to confer phenotypic resistance. Lim et al. 16 recently showed that 59% of S. pneumoniae with a levofloxacin MIC of 2 mg/l had a first step parc mutation, which raises concerns as to the possible evolution of fluoroquinolone resistance. In this study, 215 isolates (2.1%) showed levofloxacin MICs of 2 mg/l. Also of concern is a recent report in which the authors postulate that previous exposure to a fluoroquinolone may result in the development of resistance upon treatment with an additional fluoroquinolone. 17 As it is a longitudinal study, PRO- TEKT US should be able to track the emergence of fluoroquinolone resistance in pneumococci. Of concern is the increasing number of isolates in the USA with resistance to multiple antimicrobials, 8,18 as this may limit empirical treatment options for CARTIs. A recent study postulated that 41% of S. pneumoniae isolates from the Active Bacterial Core surveillance sites in the USA will be resistant to both penicillin and erythromycin by July 2004. 19 In the PROTEKT US study, b-lactam susceptibility was reduced among macrolide-resistant isolates, with only 13.5% of such isolates remaining susceptible to penicillin. Among penicillin-resistant isolates, only 19.4% were susceptible to erythromycin. Telithromycin and linezolid were the only antimicrobials in the panel that remained active against >99% of isolates. The S. pyogenes isolates collected from across the USA in 2001 2002 remained susceptible to penicillin. Susceptibility All isolates (n = 3296) b-lactamase negative (n = 2391) b-lactamase positive (n = 905) MIC 90 MIC 90 MIC 90 Ampicillin >4 <_ 0.5 >4 71.1 1 <_ 0.5 >4 98.1 >4 2 >4 0.0 Co-amoxiclav 1 <_ 0.5 >4 99.8 1 <_ 0.5 >4 99.7 2 <_ 0.5 >4 99.9 Cefprozil >8 <_ 0.12 >8 81.3 >8 <_ 0.12 >8 89.3 >8 <_ 0.12 >8 59.9 Cefuroxime 4 <_ 0.12 >16 96.3 4 <_ 0.12 >16 96.5 4 <_ 0.12 >16 95.9 Cefotaxime <_ 0.12 <_ 0.12 1 100 <_ 0.12 <_ 0.12 1 100 <_ 0.12 <_ 0.12 1 100 Clarithromycin 16 <_ 0.25 >256 80.2 16 <_ 0.25 >256 81.8 16 <_ 0.25 128 75.8 Azithromycin 2 <_ 0.12 >256 98.9 2 <_ 0.12 >256 98.8 2 <_ 0.12 128 99.2 Telithromycin 4 <_ 0.12 >16 96.3 4 <_ 0.12 >16 95.9 4 <_ 0.12 >16 97.3 Levofloxacin <_ 0.06 <_ 0.06 32 99.7 <_ 0.06 <_ 0.06 32 99.6 <_ 0.06 <_ 0.06 4 99.9 Gatifloxacin <_ 0.03 <_ 0.03 4 99.7 <_ 0.03 <_ 0.03 4 99.6 <_ 0.03 <_ 0.03 2 99.9 Moxifloxacin <_ 0.03 <_ 0.03 >4 99.7 <_ 0.03 <_ 0.03 >4 99.6 0.06 <_ 0.03 2 99.9 Tetracycline 0.5 <_ 0.06 >8 99.3 0.5 <_ 0.06 >8 99.6 0.5 <_ 0.06 >8 98.5 i13

S. D. Brown and M. J. Rybak was reduced to the macrolides irrespective of the PROTEKT US region of origin. Overall, 5.7% of isolates were resistant to erythromycin (0.3% intermediate), which is similar to the findings of Critchley et al. 20 who reported 6.2% of S. pyogenes isolates in the USA in 1999 as azithromycin resistant. Other countries have reported a higher prevalence of macrolide resistance (e.g. 28.0% in Hong Kong 21 ), indicating that macrolide resistance in S. pyogenes could increase further in the USA. With the exception of linezolid, the MIC 90 of all antimicrobials tested was <_1 mg/l for S. pyogenes; that of telithromycin was 0.03 mg/l. The telithromycin MIC for 12 isolates was >_4 mg/l; these isolates were all macrolide and clindamycin resistant (MICs >256 mg/l and >_2 mg/l, respectively) suggestive of the MLS B resistance phenotype. Previous studies have shown that the activity of ketolides may be lower against S. pyogenes isolates that carry the MLS B resistance gene, erm(b). 22 Recent evidence suggests that the prevalence of b-lactamaseproducing H. influenzae isolates may be decreasing in the USA. 7,8 The PROTEKT US data from 2001 2002 suggest a continuation of this trend. A small number of BLNAR isolates continues to be present in the USA; Thornsberry et al. 7 reported a prevalence of 0.3% (4/1152) in 1998 1999 and 0.2% (4/1934) in 1999 2000, similar to the 0.4% identified in this study. Such resistance is probably mediated by alteration of the penicillinbinding proteins. With the exception of ampicillin, the MIC 90 of the antimicrobials tested was unaffected by the b-lactamase status of the isolates; values for co-amoxiclav, cefuroxime, cefotaxime, azithromycin, telithromycin, levofloxacin, gatifloxacin, moxifloxacin and tetracycline were all <_ 4 mg/l. Ten H. influenzae isolates were identified with levofloxacin, moxifloxacin and gatifloxacin MICs above the NCCLS susceptibility breakpoints; all were isolates from New York State. Twelve such isolates, also from New York State, were isolated in the first year of the PROTEKT US programme (Aventis, data on file), but they have not been reported by others. 7,8,23,24 These isolates give cause for concern, and warrant further investigation, especially as a treatment failure with levofloxacin for H. influenzae pneumonia has recently been reported in Spain. 25 In conclusion, results from the 2001 2002 season of PRO- TEKT US indicate that antimicrobial resistance continues to be a problem among CARTI-causing pathogens in the USA and warrants continued monitoring in the future through longitudinal surveillance studies, such as PROTEKT US. The ketolide telithromycin continues to show high activity against common CARTI pathogens, including those that have developed resistance to the b-lactams and macrolides. Acknowledgements We gratefully acknowledge the centres that provided isolates for the second year of the PROTEKT US study. We also thank the researchers of CMI (Wilsonville, OR, USA) for their microbiological analysis of the isolates. Data analysis was carried out by Micron Research Ltd (Upwell, Cambridgeshire, UK). The PRO- TEKT US study is supported by Aventis. Transparency declarations S.D.B. has received funds for research from Aventis Pharmaceuticals and has served on an advisory board for telithromycin. M.J.R. has received a consulting fee for a meeting sponsored by Aventis Pharmaceuticals and an honorarium for speaking at a symposium that received educational support from Aventis. References 1. Adam, D. (2002). Global antibiotic resistance in Streptococcus pneumoniae. Journal of Antimicrobial Chemotherapy 50, Topic T1, 1 5. 2. National Committee for Clinical Laboratory Standards. (2000). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically Fifth Edition: Approved Standard M7-A5. NCCLS, Villanova, PA, USA. 3. National Committee for Clinical Laboratory Standards. (2002). Performance Standards for Antimicrobial Sensitivity Testing Twelfth Edition. Supplemental Tables M100-S12. NCCLS, Villanova, PA, USA. 4. National Committee for Clinical Laboratory Standards. (2004). Performance Standards for Antimicrobial Sensitivity Testing Fourteenth Edition. Supplemental Tables M100-S14. NCCLS, Villanova, PA, USA. 5. Doern, G. V., Heilmann, K. P., Huynh, H. K. et al. (2001). Antimicrobial resistance among clinical isolates of Streptococcus pneumoniae in the United States during 1999-2000, including a comparison of resistance rates since 1994-1995. Antimicrobial Agents and Chemotherapy 45, 1721 9. 6. Karlowsky, J. A., Thornsberry, C., Jones, M. E. et al. (2003). Factors associated with relative rates of antimicrobial resistance among Streptococcus pneumoniae in the United States: results from the TRUST Surveillance Program (1998-2002). Clinical Infectious Diseases 36, 963 70. 7. Thornsberry, C., Sahm, D. F., Kelly, L. J. et al. (2002). Regional trends in antimicrobial resistance among clinical isolates of Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis in the United States: results from the TRUST Surveillance Program, 1999-2000. Clinical Infectious Diseases 34, Suppl.1, S4 S16. 8. Karlowsky, J. A., Draghi, D. C., Thornsberry, C. et al. (2002). Antimicrobial susceptibilities of Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis isolated in two successive respiratory seasons in the US. International Journal of Antimicrobial Agents 20, 76 85. 9. Jacobs, M. R., Bajaksouzian, S., Zilles, A. et al. (1999). Susceptibilities of Streptococcus pneumoniae and Haemophilus influenzae to 10 oral antimicrobial agents based on pharmacodynamic parameters: 1997 US surveillance study. Antimicrobial Agents and Chemotherapy 43, 1901 8. 10. Doern, G. V., Brueggemann, A. B., Huynh, H. et al. (1999). Antimicrobial resistance with Streptococcus pneumoniae in the United States, 1997-98. Emerging Infectious Diseases 5, 757 65. 11. Jones, M. E., Karlowsky, J. A., Blosser-Middleton, R. et al. (2002). Longitudinal assessment of antipneumococcal susceptibility in the United States. Antimicrobial Agents and Chemotherapy 46, 2651 5. 12. Farrell, D. J. & Jenkins, S. G. (2004). Distribution across the USA of macrolide resistance and macrolide resistance mechanisms among Streptococcus pneumoniae isolates collected from patients with respiratory tract infections: PROTEKT US 2001-2002. Journal of Antimicrobial Chemotherapy 54, Suppl. S1, i17 22. 13. Doern, G. V. & Brown, S. D. (2004). Antimicrobial susceptibility among community-acquired respiratory tract pathogens in the USA: data from PROTEKT US 2000-01. Journal of Infection 48, 56 65. 14. Brueggemann, A. B., Coffman, S. L., Rhomberg, P. et al. (2002). Fluoroquinolone resistance in Streptococcus pneumoniae in the United States since 1994-1995. Antimicrobial Agents and Chemotherapy 46, 680 8. 15. Ruiz, J. (2003). Mechanisms of resistance to quinolones: target alterations, decreased accumulation and DNA gyrase protection. Journal of Antimicrobial Chemotherapy 51, 1109 17. i14

Susceptibility of RTI pathogens in 2001 2002 in the USA 16. Lim, S., Bast, D., McGeer, A. et al. (2003). Antimicrobial susceptibility breakpoints and first-step parc mutations in Streptococcus pneumoniae: redefining fluoroquinolone resistance. Emerging Infectious Diseases 9, 833 7. 17. d Escrivan, T., Roussel-Delvallez, M., Alfandari, S. et al. (2003). High level of resistance to levofloxacin of Streptococcus pneumoniae following fluoroquinolone therapy. Infection 31, 366. 18. Whitney, C. G., Farley, M. M., Hadler, J. et al. (2000). Increasing prevalence of multidrug-resistant Streptococcus pneumoniae in the United States. New England Journal of Medicine 343, 1917 24. 19. McCormick, A. W., Whitney, C. G., Farley, M. M. et al. (2003). Geographic diversity and temporal trends of antimicrobial resistance in Streptococcus pneumoniae in the United States. Nature Medicine 9, 424 30. 20. Critchley, I. A., Sahm, D. F., Thornsberry, C. et al. (2002). Antimicrobial susceptibilities of Streptococcus pyogenes isolated from respiratory and skin and soft tissue infections: United States LIBRA surveillance data from 1999. Diagnostic Microbiology and Infectious Disease 42, 129 35. 21. Cantón, R., Loza, E., Morosini, M. I. et al. (2002). Antimicrobial resistance amongst isolates of Streptococcus pyogenes and Staphylococcus aureus in the PROTEKT antimicrobial surveillance programme during 1999 2000. Journal of Antimicrobial Chemotherapy 50, Suppl. S1, 9 24. 22. Giovanetti, E., Montanari, M. P., Marchetti, F. et al. (2000). In vitro activity of ketolides telithromycin and HMR 3004 against Italian isolates of Streptococcus pyogenes and Streptococcus pneumoniae with different erythromycin susceptibility. Journal of Antimicrobial Chemotherapy 46, 905 8. 23. Doern, G. V., Jones, R. N., Pfaller, M. A. et al. (1999). Haemophilus influenzae and Moraxella catarrhalis from patients with community-acquired respiratory tract infections: antimicrobial susceptibility patterns from the SENTRY antimicrobial surveillance program (United States and Canada, 1997). Antimicrobial Agents and Chemotherapy 43, 385 9. 24. Thornsberry, C., Ogilvie, P. T., Holley, H. P. et al. (1999). Survey of susceptibilities of Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis isolates to 26 antimicrobial agents: a prospective US study. Antimicrobial Agents and Chemotherapy 43, 2612 23. 25. Bastida, T., Pérez-Vázquez, M., Campos, J. et al. (2003). Levofloxacin treatment failure in Haemophilus influenzae pneumonia. Emerging Infectious Diseases 9, 1475 8. i15