- Email: [email protected]
Antibiotic resistance among recent clinical isolates of Haemophilus influenzae in Japanese children
Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
Antibiotic resistance among recent clinical isolates of Haemophilus influenzae in Japanese children Kiyofumi Ohkusua,*, Akira Nakamurab, Kyoko Sawadaa a
Institution:Division of Clinical Laboratory, Chiba Children’s Hospital, 579-1 Heta cho, Midori-ku, Chiba–City 266-0007 Japan b Division of Infectious Diseases, Chiba Children’s Hospital, 579-1 Heta cho, Midori-ku, Chiba–City 266-0007 Japan Received 27 October 1999; accepted 2 December 1999
Abstract From January 1997 to July 1999, a total of 867 isolates of Haemophilus influenzae were recovered in the microbiology laboratory of Chiba Children’s Hospital. The overall prevalence of -lactamase production was 12.8%. Ampicillin-MICs for all of the 111 -lactamaseproducing isolates was ⱖ4 g/ml. A total of 26 -lactamase-negative isolates (3.4% of all -lactamase-negative isolates and 3.0% of all isolates) were found to be resistant to ampicillin. The prevalence of -lactamase negative ampicillin-resistant strains (BLNAR) increased remarkably to 8.9% during the last 7-month period. It is noteworthy that the MICs not only of penicillins but also of cephems for BLNAR were significantly higher than those for ampicillin-susceptible isolates. Eight -lactamase-producing isolates of H. influenzae (7.2% of all -lactamase-producing isolates) were resistant to amoxicillin-clavulanate (AMPC/CVA). Consequently, the overall resistance to ampicillin was 15.8%, and that to AMPC/CVA was 3.0%. The results of this study corroborate the findings of previous investigators in the US (Doern et al., 1997) regarding the emergence of BLNAR and -lactamase-producing AMPC/CVA-resistant strains (BLPACR) of H. influenzae. Continued monitoring of susceptibility trends will be required to guide appropriate chemotherapy. © 2000 Elsevier Science Inc. All rights reserved.
1. Introduction Haemophilus influenzae is one of the most common organisms causing meningitis, epiglottitis, acute otitis media, and lower respiratory tract infections in children. Since the first cases of documented -lactamase-mediated ampicillin resistance among H. influenzae in the US in 1974, the prevalence of -lactamase-producing strains has steadily increased and now exists at a level of approximately 36% in the US (Doern et al., 1997; Jones et al., 1997). In Japan, approximately 17% of the strains recovered from pediatric patients were -lactamase-producing ampicillin resistant (Nakamura et al., 1995; Ishiwada et al., 1998). Of particular concern in the results of the surveillance study in the US was the recovery of ampicillin resistance among -lactamase-negative strains and amoxicillin-clavulanate resistance among -lactamase-producing strains (Doern et al., 1997). The in vitro activities of some cephems were also diminished against these strains (Barry et al., 1993). There* Corresponding author. Tel.: ⫹81-43-292-2111; fax: ⫹81-43-2923815. E-mail address: [email protected] (K. Ohkusu).
fore, if the rates of these strains continue to increase, the empiric management of H. influenzae infections would become complicated and in vitro susceptibility testing would need to be considered. With this in mind, we evaluated antimicrobial susceptibility and monitored the emerging resistance among clinical isolates of H. influenzae recovered from pediatric patients.
2. Materials and methods A total of 867 isolates of H. influenzae were recovered between January 1997 and July 1999 in the microbiology laboratory of Chiba Children’s Hospital, which is a 200-bed scaled tertiary-care hospital. Clinical isolates were identified as H. influenzae by standard methodology (Campos, 1995). Subcultures were performed with chocolate agar plates (Becton–Dickinson, Tokyo, Japan) and incubated at 35°C in 5% CO2 for 16 to 18 h. Minimum inhibitory concentrations (MICs) were determined as described according to the Japan Society of Chemotherapy Standards by broth microdilution procedure (100 l total volume per well; final inoculum concentration, approximately 5 ⫻ 105
0732-8893/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 2 - 8 8 9 3 ( 9 9 ) 0 0 1 4 7 - 9
250
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
Table 1 Source of 867 clinical isolates of H. influenzae by patient sex, age, and specimen Characteristics and description Gender Male Female Age (yr) 0 1 2 3 4 5 6 7 8 9 ⱖ10 Specimen type Sputum Nasal/Nasopharynx Otorrhea Eye swab Blood/CSFb Throat Other
Total no. (%) of isolates
% of isolates -Lactamasepositive
523 (60.3) 344 (39.7)
11.9 14.2
3.6 2.0
75 (8.7) 125 (14.4) 116 (13.4) 107 (12.3) 109 (12.6) 100 (11.5) 60 (6.9) 60 (6.9) 29 (3.3) 37 (4.3) 49 (5.7)
18.7 12.8 9.5 10.3 13.8 13.0 20.0 13.3 3.4 13.5 10.2
1.3 6.4 0.9 1.9 4.6 2.0 5.0 3.3 0 5.4 0
472 (54.4) 304 (35.1) 46 (5.3) 31 (3.6) 6 (0.7) 5 (0.6) 3 (0.3)
13.8 10.9 4.3 19.4 0 80.0a 33.3
3.4 2.6 2.2 0 0 20.0 0
cially (Eiken, Tokyo, Japan) and the following 10 antimicrobials were tested: ampicillin, amoxicillin-clavulanate (ACV, 2:1 ratio of amoxicillin to clavulanate), cefaclor, cefixime, cefotaxime, ceftriaxone, imipenem, norfloxacin, erythromycin, and chloramphenicol. H. influenzae ATCC™ 49247 and ATCC™ 49766 were used as quality-controls (QC). -Lactamase production was assessed with the nitrocefin disk assay (Cefinase™; Becton–Dickinson, MD, USA). Serotyping was performed by using the slide agglutination procedure with the Phadebact™ Hemophilus Test kit (Pharmacia, Sweden) and type b antiserum (bı´oMe´riex, France). Statistical analysis of data was carried out by using the 2 method.
% of isolates BLNAR
3. Results Table 1 summarizes the sources of H. influenzae by patient sex, age, and specimen type. There was no significant difference in the prevalence of -lactamase production and BLNAR isolates between isolates from males and females. More than 70% of the isolates were obtained from patients ages ⱕ 5 years old and there was also no significant difference in the prevalence of -lactamase production and BLNAR isolates by patient age. The majority of the isolates (89.5%) were recovered from respiratory tract such as sputum, nasal, or nasopharynx specimen. The overall prevalence of -lactamase producer was 12.8%. Table 2 shows the results of MICs determinations with ten selected antimicrobial agents comparing -lactamase-negative and -positive strains. Although the QC results were within recommended the National Committee for Clinical Laboratory Standards (NCCLS) limits and breakpoints were used to interpret the results (NCCLS, 1998), Mueller–Hinton broth was used rather than the NCCLS recommended HTM broth. On the basis of current NCCLS breakpoints (NCCLS, 1998), the overall percentages of isolates found to be resistant or intermediate to the following
a Indicates statistically significant difference (P ⬍ 0.05 by chi-square test) in prevalence of -lactamase-positive or BLNAR isolates. b CSF: cerebrospinal fluid.
colony forming U/ml). The trays were incubated at 35°C in ambient air for 18 to 24 h before determining the results. Inoculum size was confirmed by performing viable organism counts. Susceptibility testing medium broth was prepared by using cation-adjusted Mueller–Hinton broth (Becton–Dickinson, MD, USA) with 3% lysed horse blood, 15 mg of NAD per ml, and 5 mg of yeast extract per ml (Eiken, Tokyo, Japan). Antibiotic panels were obtained commerTable 2 MICs of 10 antimicrobials for recent clinical isolates of H. influenzae Antimicrobial agent
Ampicillin Amoxicillin-clavulanatea Cefaclor Cefixime Cefotaxime Ceftriaxone Imipenem Norfloxacin Erythromycin Chloramphenicol a
Total no. of isolates
867 867 732 732 867 867 732 867 867 867
2 : 1 ratio of amoxicillin to clavulanate.
MIC (g/ml)
-Lactamase-negative isolates
-Lactamase-positive isolates
No. of isolates
50%
90%
Range
No. of isolates
50%
90%
Range
756 756 638 638 756 756 638 756 756 756
0.25 0.75 2 0.06 0.03 ⱕ0.015 1 ⱕ0.06 4 0.5
1 3 ⬎16 0.13 0.25 0.03 2 0.13 8 1
ⱕ0.13⬃16 ⱕ0.19⬃24 ⱕ0.25⬃⬎16 ⱕ0.03⬃4 ⱕ0.015⬃4 ⱕ0.015⬃1 ⱕ0.06⬃8 ⱕ0.06⬃1 ⱕ0.13⬃⬎16 ⱕ0.13⬃8
111 111 94 94 111 111 94 111 111 111
⬎16 1.5 8 0.06 0.03 ⱕ0.015 1 ⱕ0.06 4 1
⬎16 6 ⬎16 0.25 0.25 0.03 4 0.13 8 16
4⬃⬎16 0.38⬃12 1⬃⬎16 ⱕ0.03⬃⬎4 ⱕ0.015⬃0.25 ⱕ0.015⬃0.25 0.13⬃8 ⱕ0.06⬃4 2⬃16 0.25⬃16
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
agents were: 13.0% to cefaclor, 1.5% to cefixime, 0.3% to cefotaxime, 0% to ceftriaxone, 1.8% to imipenem, and 4.2% to chloramphenicol. The MICs of erythromycin at which 50% of isolates are inhibited (MIC50) and MIC90 were 4 and 8 g/ml, respectively. The MIC50 and MIC90 of norfloxacin were ⱕ0.06 and 0.13 g/ml, respectively. The norfloxacin MICs of one isolate was 4 g/ml. The following chloramphenicol MICs were obtained with -lactamase-negative and -positive strains, respectively: MIC50, 0.5 and 1 g/ml; MIC90, 1 and 16 g/ml; and the number of strains and the percentage of MICs ⱖ8 g/ml, 1 (0.1%) and 33 (29.7%). The activity of chloramphenicol against -lactamase-positive isolates was significantly diminished compared with that against -lactamase-negative isolates. The activity of cefaclor against -lactamase-positive isolates was slightly diminished compared with that against -lactamase-negative isolates. On the basis of a comparison of the MIC50 and MIC90, -lactamase production did not influence the activity of cefixime, cefotaxime, ceftriaxone, or norfloxacin. A total of 26 -lactamase-negative isolates (3.4% of all -lactamase-negative isolates and 3.0% of all isolates) was found to be BLNAR. The actual distribution of MICs for BLNAR isolates were 4 g/ml, 18; 8 g/ml, 6; and 16 g/ml, two strains. Isolation of BLNAR with high-level resistance (ampicillin-MICs ⱖ 8 g/ml) has also been gradually increasing (1997, 0; 1998, 3; 1999 [until July], five strains). The number of -lactamase-producing isolates for which the ampicillin-MICs were 4, 8, 16, and ⬎16 g/ml were four (3.6%), nine (8.1%), six (5.4%), and 92 strains (82.9%), respectively. Eight -lactamase-positive isolates (7.2% of all -lactamase-producers) were BLPACR. For all eight BLPACR strains, ACV-MICs were 8/4 g/ml, and ampicillin-MICs were ⱖ64 g/ml. Determination of ampicillin- and ACV-MICs for BLNAR or BLPACR were repeated twice, and was never varied by more than ⫾ 1 twofold concentration (data not shown). The -lactamase production was also confirmed by repeat testing. Thus, the overall resistance to ampicillin was 15.8%, and that to ACV was 3.0%. None of the 26 BLNAR and 8 BLPACR isolate were serologically typeable strains, and most of them were recovered from the respiratory tract. The distribution of ACV-MICs for ampicillin-nonsusceptible isolates (ampicillin-MICs ⱖ 2 g/ml; n ⫽ 43) were 2/1 g/ml, 10; 4/2 g/ml, 15; 8/4 g/ml, 13; and 16/8 g/ml, five strains, whereas the modal ampicillin- and ACV-MICs for -lactamase-negative ampicillin-susceptible isolates (ampicillin-MICs ⱕ 1 g/ml) were 0.25 and 0.5/ 0.25 g/ml, respectively. The modal MICs of cefaclor, cefixime, cefotaxime, ceftriaxone, and imipenem for ampicillin-nonsusceptible isolates were ⬎16, 0.06 to 0.13, 0.25, 0.03, and 4 to 8 g/ml, respectively. By comparison, the modal MICs obtained with the same agents for ampicillinsusceptible isolates were two to four, ⱕ0.03, 0.03, ⱕ0.015, and 0.5 to 1 g/ml, respectively. Figs. 1a and b illustrate the distribution of antimicrobial susceptibilities of H. influenzae by -lactamase production
251
and ampicillin-MICs levels. The sensitivity distribution showed a trend toward higher MICs of cephalosporins for -lactamase-negative isolates and correspondingly the rate of isolates with higher ampicillin-MICs also increased. The MICs distribution of cephems for ampicillin-nonsusceptible isolates shifted from 2 to 6 dilutions higher than those for other -lactamase-negative isolates. Fig. 2 shows comparison of cumulative MICs of four cephems for ampicillinnonsusceptible isolates and ampicillin-susceptible strains. Based on the comparison of MICs distribution, ceftriaxone was the most active against H. influenzae in the cephems examined in this study. MICs of ceftriaxone tended to be from 2- to fourfold lower than that of cefotaxime for both ampicillin-nonsusceptible isolates and -lactamase producers. However, both cefotaxime- and ceftriaxone-MICs for ampicillin-nonsusceptible isolates shifted toward higher MICs just as the other cephems did.
4. Discussion H. influenzae causes a variety of community-acquired bacterial infections, especially in children. Nontypeable strains of H. influenzae are the most frequent pathogens causing bronchopneumonia, bronchitis, sinusitis, and acute otitis media. In Japan, H. influenzae serotype b (Hib) still remains the most significant cause of bacterial meningitis, septic arthritis, and epiglottitis in children, because Hib vaccines are not currently available (Uehara et al., 1998). The most common mechanism of ampicillin-resistance with H. influenzae is the production either of two -lactamase, TEM-1 or ROB-1 (Scriver et al., 1994). The results of this study indicated that the overall prevalence of -lactamase production among clinical isolates of H. influenzae was much lower than that reported in the US study (12.8% versus 36.0% for 1994 to 1995) (Jones et al., 1997). Recently, in the SENTRY antimicrobial surveillance program of 837 respiratory tract isolates, the percentage of -lactamase-producing H. influenzae was 33.5% in North America (Doern et al., 1999). In Japan, after comparing an earlier study (Nakamura et al., 1995), it seems that the prevalence of -lactamase-mediated ampicillin-resistance of H. influenzae is not increasing among children. The prevalence of BLPACR’s was similar to that reported in the US (0.8% versus 1.1, 0.2%) (Doern et al., 1997, 1999). However, the prevalence of BLNAR’s found in this study was somewhat higher than that reported in the US (3.0% versus 0.8, 0.1%). In addition, the finding in these surveillance studies have not confirmed the increasing incidents of BLNAR’s in the US (0.8% in 1994 –1995, 0.1% in 1997) (Doern et al., 1997, 1999). Of special interest to us is the prevalence of BLNAR’s increased remarkably to 8.9% during the last 7-month period. The ampicillin-MICs of BLNAR’s have also been gradually increasing. Furthermore, the in vitro activities were significantly diminished against these
252
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
Fig. 1. Frequency distribution of antimicrobial MICs values for H. influenzae isolates, by -lactamase production and ampicillin MICs-levels.
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
253
Fig. 2. Comparison of cumulative MICs four cephems for ampicillin-nonsusceptible isolates (Solid line; ampicillin MICs ⱖ 2 g/ml) and ampicillinsusceptible isolates (Dotted line; ampicillin MICs ⱕ 1 /ml) of H. influenzae.
strains, not only ACV, the -lactamase inhibitor combination, but also other penicillins and cephems. These BLNAR isolates seem to have altered penicillinbinding proteins (PBPs) with decreased affinity or diminished permeability for ampicillin and other -lactams. The characteristics of phenotypic expressions that were observed in this study were consistent with this mechanism. Claioux et al. (1992) reported that changes in PBPs-3A and -3B are a common mechanism that results in a significant level of non-enzymatic resistance to -lactams in H. influenzae regardless of serotype, origin of the isolates, or geographic distribution. Moreover, the level of resistance is related to the extent of PBP modifications. On the other hand, several studies have reported a strain with PBPs-4 and -5 as having a decreased affinity for -lactams (Mendelman et al., 1984; Serfass et al., 1986; Mendelman et al., 1990). On the basis of these previous studies, for a more reliable and effective interpretation of BLNAR, the genes responsible for PBP modification should be clarified as the same as demonstrated in Streptococcus pneumoniae such as pbp2x, pbp1a, and pbp2b (Ubukata et al., 1996; Asahi et al., 1998) The results of this study corroborate the finding in the aforementioned investigation (Doern et al., 1999) regarding the rate of resistant to cefaclor (13.0% versus 12.8%). However, it is interesting that chloramphenicol-resistance of our study was much higher than that reported in the US (4.2% versus 0.7%). Two prior multicenter studies in the US (Doern et al., 1997; Jorgensen et al., 1990) reported chloramphenicol-resistant strains were also uncommon (0.2% and 0%, respectively). With the notable observations in this study, the majority (92%) of the intermediate or resistant isolates for chloramphenicol was -lactamase producers. Thus, chloramphenicol-resistance was more common among -lactamase-positive isolates than -lactamase-negative isolates in Japan (29.7% versus 0.4%). The fluoroquinolones exhibited good activity against both BLNAR’s and BLPACR’s. However, as observed in
this study, fluoroquinolone-resistant H. influenzae strains have been reported by several investigators (Georgiou et al., 1996; Bootsma et al., 1997; Vila et al., 1999). According to these studies, the mechanisms of fluoroquinolone-resistance are mainly due to chromosomal mutations in the gyrA and parC genes, which encode the A subunits of the DNA gyrase and topoisomerase IV, respectively. It will be required to examine the MICs of fluoroquinolone or these gene mutations for isolates recovered from patients with recurrent H. influenzae infections after or during fluoroquinolone therapy. It will be also important to monitor the susceptibility of H. influenzae as their frequent use. Indeed, the actual clinical implications of elevated -lactam MICs for BLNAR or BLPACR strains is poorly understood. One case report has indicated that BLNAR was recovered from the cerebrospinal fluid and its altered PBPs for -lactam were associated with cefuroxime treatment failure (Mendelman et al., 1990). If the frequency of BLNAR’s continue to increase, the incidence of clinical treatment failure cases would be increase in the near future. In our study, none of the BLNAR’s or BLPACR’s was serologically typeable strains and almost all of them were from respiratory tract infections. If BLNAR’s or BLPACR’s emerge among Hib isolates, clinical problems would become much more critical, because Hib vaccines are not currently used in Japan. In conclusion, it is important for laboratory strategies to examine routinely not only -lactamase production, but also to carry out in vitro susceptibility testing, especially MIC determination. In addition, continued monitoring of susceptibility trends will be required to guide appropriate antimicrobial chemotherapy. References Asahi, Y. & Ubukata, K. (1998). Association of a Thr-371 substitution in a conserved amino acid motif of penicillin-binding protein 1A with
254
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
penicillin resistance of Streptococcus pneumoniae. Antimicrob Agents Chemother 42, 2267–2273. Barry, A. L., Fuchs, P. C., & Pfaller, M. A. (1993). Susceptibilities of -lactamase-producing and -nonproducing ampicillin-resistant strains of Haemophilus influenzae toceftibuten, cefaclor, cefuroxime, cefixime, cefotaxime, and amoxicillin-clavulanic acid. Antimicrob Agents Chemother 37, 14 –18. Bootsma, H. J., Troeltra, A., Veen–Rutgers, A., Mooi, F. R., Neeling, A. J., & Overbeek, B. P. (1997). Isolation and characterization of a ciprofloxacin-resistant isolate of Haemophilus influenzae from The Netherlands. J Antimicrob Chemother 39, 292–293. Campos, J. M. (1995). Haemophilus. In P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, & R. H. Yolken (Eds.). Manual of Clinical Microbiology, 6th edition (pp. 556 –565). Washington, D.C.: ASM Press. Clairoux, N., Picard, M., Brochu, A., Rousseau, N., Gourde, P., Beauchamp, D., Parr, T. R., Bergeron, M. G., & Malouin, F. (1992). Molecular basis of the non- -lactamase-mediated resistance to -lactam antibiotics in strains of Haemophilus influenzae isolated in Canada. Antimicrob Agents Chemother 36, 1504 –1513. Doern, G. V., Brueggemann, A. B., Pierce, G., Holley, H. P., & Rauch, A. (1997). Antibiotic resistance among clinical isotates of Haemophilus influenzae in the United States in 1994 and 1995 and detection of -lactamase-positive strains resistant to amoxicillin-clavulanate: results of a national multicenter surveillance study. Antimicrob Agents Chemother 41, 292–297. Doern, G. V., Jones, R. N., Pfaller, M. A., Kugler, K., & the SENTRY participants group (1999). Haemophilus influenzae and Moraxella catarrhalis from patients with community-acquired respiratory tract infections: antimicrobial susceptibility patterns from the SENTRY antimicrobial surveillance program (US and Canada, 1997). Antimicrob Agents Chemother 43, 385–389. Georgiou, M., Munoz, R., Roman, F., Canton, R., Gomez–Lus, R., Campos, J., & Campa, A. G. (1996). Ciprofloxacin-resistant Haemophilus influenzae strains possess mutations in analogous positions of gyrA and. parC Antimicrob Agents Chemother 40, 1741–1744. Ishiwada, N., Kuroki, H., Ishikawa, N., Sugimoto, K., Koori, Y., Suruga, Y., Uehara, S., & Niimi, H. (1998). Characteristics of -lactamaseproducing and amoxicillin-clavulanate-resistant strains of Haemophilus influenzae isolated from pediatric patients. J Infect Chemother 4, 112– 115. Jones, R. N., Jacobs, M. R., Washington, J. A., & Pfaller, M. A. (1997). A 1994 –95 survey of Haemophilus influenzae susceptibility to ten orally
administered agents A 187 clinical laboratory center sample in the United States. Diagn Microbiol Infect Dis 27, 75– 83. Jorgensen, J. H., Doern, G. V., Maher, L. A., Howell, A. W., & Redding, J. S. (1990). Antimicrobial resistance among respiratory isolates of Haemophilus influenzae, Moraxella catarrhalis, and Streptococcus pneumoniae in the US. Antimicrob Agents Chemother 34, 2075–2080. Mendelman, P. M., Chaffin, D. O., Stull, T. L., Rubens, C. E., Mack, K. D., & Smith, A. L. (1984). Characterization of non--lactamase-mediated ampicillin resistance in Haemophilus influenzae. Antimicrob Agents Chemother 26, 235–244. Mendelman, P. M., Chaffin, D. O., Krilov, L. R., Kalaitzoglou, G., Serfass, D. A., Onay, O., Wiley, E. A., Overturf, G. D., & Rubin, L. G. (1990). Cefuroxime treatment failure of nontypable Haemophilus influenzae meningitis associated with alteration of penicillin-binding proteins. J Infect Dis 162, 1118 –1123. Nakamura, A., Ishikawa, N., Suzuki, H., Kurosaki, T., Terashima, I., Uehara, S., & Niimi, H. (1995). Clinical bacteriology of Haemophilus influenzae isolated from respiratory tract infections in children. J Jpn Pediatr Soc 99, 1234 –1241. National Committee for Clinical Laboratory Standards (NCCLS) (1998). Performance standards for antimicrobial susceptibility testing. Supplement tables, M100-S8. Wayne, PA: NCCLS. Scriver, S. R., Walmsley, S. L., Kau, C. L., Hoban, D. J., Brunton, J., McGeer, A., Moore, T. C., Witwicki, E., Canadian Haemophilus Study Groop, & Low, D. E. (1994). Determination of antimicrobial susceptibilities of Canadian isolates of Haemophilus influenzae and characterization of their -lactamases. Antimicrob Agents Chemother 38, 1678 –1680. Serfass, D. A., Mendelman, P. M., Chaffin, D. O., & Needman, C. A. (1986). Ampicillin resistance and penicillin-binding proteins of Haemophilus influenzae. J Gen Microbiol 132, 2855–2861. Ubukata, K., Asahi, Y., Yamane, A., & Konno, M. (1996). Combinational detection of autolysin and penicillin-binding protein 2B genes of Streptococcus pneumoniae by PCR. J Clin Microbiol 34, 592–596. Uehara, S., Kamiya, H., Togashi, T., Kato, T., Shiraki, K., & Morishima, T. (1998). Epidemiology of Haemophilus influenzae meningitis among infants and children in Japan comparison with incidences of bacterial meningitis. J Jpn Pediatr Soc 102, 656 – 665. Vila, J., Ruiz, J., Sanchez, F., Navarro, F., Mirelis, B., Anta, M. E. J., & Prats, G. (1999). Increase in quinolone resistance in a Haemophilus influenzae strain isolated from a patient with recurrent respiratory infections treated with ofloxacin. Antimicrob Agents Chemother 43, 161–162.
Antibiotic resistance among recent clinical isolates of Haemophilus influenzae in Japanese children Kiyofumi Ohkusua,*, Akira Nakamurab, Kyoko Sawadaa a
Institution:Division of Clinical Laboratory, Chiba Children’s Hospital, 579-1 Heta cho, Midori-ku, Chiba–City 266-0007 Japan b Division of Infectious Diseases, Chiba Children’s Hospital, 579-1 Heta cho, Midori-ku, Chiba–City 266-0007 Japan Received 27 October 1999; accepted 2 December 1999
Abstract From January 1997 to July 1999, a total of 867 isolates of Haemophilus influenzae were recovered in the microbiology laboratory of Chiba Children’s Hospital. The overall prevalence of -lactamase production was 12.8%. Ampicillin-MICs for all of the 111 -lactamaseproducing isolates was ⱖ4 g/ml. A total of 26 -lactamase-negative isolates (3.4% of all -lactamase-negative isolates and 3.0% of all isolates) were found to be resistant to ampicillin. The prevalence of -lactamase negative ampicillin-resistant strains (BLNAR) increased remarkably to 8.9% during the last 7-month period. It is noteworthy that the MICs not only of penicillins but also of cephems for BLNAR were significantly higher than those for ampicillin-susceptible isolates. Eight -lactamase-producing isolates of H. influenzae (7.2% of all -lactamase-producing isolates) were resistant to amoxicillin-clavulanate (AMPC/CVA). Consequently, the overall resistance to ampicillin was 15.8%, and that to AMPC/CVA was 3.0%. The results of this study corroborate the findings of previous investigators in the US (Doern et al., 1997) regarding the emergence of BLNAR and -lactamase-producing AMPC/CVA-resistant strains (BLPACR) of H. influenzae. Continued monitoring of susceptibility trends will be required to guide appropriate chemotherapy. © 2000 Elsevier Science Inc. All rights reserved.
1. Introduction Haemophilus influenzae is one of the most common organisms causing meningitis, epiglottitis, acute otitis media, and lower respiratory tract infections in children. Since the first cases of documented -lactamase-mediated ampicillin resistance among H. influenzae in the US in 1974, the prevalence of -lactamase-producing strains has steadily increased and now exists at a level of approximately 36% in the US (Doern et al., 1997; Jones et al., 1997). In Japan, approximately 17% of the strains recovered from pediatric patients were -lactamase-producing ampicillin resistant (Nakamura et al., 1995; Ishiwada et al., 1998). Of particular concern in the results of the surveillance study in the US was the recovery of ampicillin resistance among -lactamase-negative strains and amoxicillin-clavulanate resistance among -lactamase-producing strains (Doern et al., 1997). The in vitro activities of some cephems were also diminished against these strains (Barry et al., 1993). There* Corresponding author. Tel.: ⫹81-43-292-2111; fax: ⫹81-43-2923815. E-mail address: [email protected] (K. Ohkusu).
fore, if the rates of these strains continue to increase, the empiric management of H. influenzae infections would become complicated and in vitro susceptibility testing would need to be considered. With this in mind, we evaluated antimicrobial susceptibility and monitored the emerging resistance among clinical isolates of H. influenzae recovered from pediatric patients.
2. Materials and methods A total of 867 isolates of H. influenzae were recovered between January 1997 and July 1999 in the microbiology laboratory of Chiba Children’s Hospital, which is a 200-bed scaled tertiary-care hospital. Clinical isolates were identified as H. influenzae by standard methodology (Campos, 1995). Subcultures were performed with chocolate agar plates (Becton–Dickinson, Tokyo, Japan) and incubated at 35°C in 5% CO2 for 16 to 18 h. Minimum inhibitory concentrations (MICs) were determined as described according to the Japan Society of Chemotherapy Standards by broth microdilution procedure (100 l total volume per well; final inoculum concentration, approximately 5 ⫻ 105
0732-8893/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 2 - 8 8 9 3 ( 9 9 ) 0 0 1 4 7 - 9
250
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
Table 1 Source of 867 clinical isolates of H. influenzae by patient sex, age, and specimen Characteristics and description Gender Male Female Age (yr) 0 1 2 3 4 5 6 7 8 9 ⱖ10 Specimen type Sputum Nasal/Nasopharynx Otorrhea Eye swab Blood/CSFb Throat Other
Total no. (%) of isolates
% of isolates -Lactamasepositive
523 (60.3) 344 (39.7)
11.9 14.2
3.6 2.0
75 (8.7) 125 (14.4) 116 (13.4) 107 (12.3) 109 (12.6) 100 (11.5) 60 (6.9) 60 (6.9) 29 (3.3) 37 (4.3) 49 (5.7)
18.7 12.8 9.5 10.3 13.8 13.0 20.0 13.3 3.4 13.5 10.2
1.3 6.4 0.9 1.9 4.6 2.0 5.0 3.3 0 5.4 0
472 (54.4) 304 (35.1) 46 (5.3) 31 (3.6) 6 (0.7) 5 (0.6) 3 (0.3)
13.8 10.9 4.3 19.4 0 80.0a 33.3
3.4 2.6 2.2 0 0 20.0 0
cially (Eiken, Tokyo, Japan) and the following 10 antimicrobials were tested: ampicillin, amoxicillin-clavulanate (ACV, 2:1 ratio of amoxicillin to clavulanate), cefaclor, cefixime, cefotaxime, ceftriaxone, imipenem, norfloxacin, erythromycin, and chloramphenicol. H. influenzae ATCC™ 49247 and ATCC™ 49766 were used as quality-controls (QC). -Lactamase production was assessed with the nitrocefin disk assay (Cefinase™; Becton–Dickinson, MD, USA). Serotyping was performed by using the slide agglutination procedure with the Phadebact™ Hemophilus Test kit (Pharmacia, Sweden) and type b antiserum (bı´oMe´riex, France). Statistical analysis of data was carried out by using the 2 method.
% of isolates BLNAR
3. Results Table 1 summarizes the sources of H. influenzae by patient sex, age, and specimen type. There was no significant difference in the prevalence of -lactamase production and BLNAR isolates between isolates from males and females. More than 70% of the isolates were obtained from patients ages ⱕ 5 years old and there was also no significant difference in the prevalence of -lactamase production and BLNAR isolates by patient age. The majority of the isolates (89.5%) were recovered from respiratory tract such as sputum, nasal, or nasopharynx specimen. The overall prevalence of -lactamase producer was 12.8%. Table 2 shows the results of MICs determinations with ten selected antimicrobial agents comparing -lactamase-negative and -positive strains. Although the QC results were within recommended the National Committee for Clinical Laboratory Standards (NCCLS) limits and breakpoints were used to interpret the results (NCCLS, 1998), Mueller–Hinton broth was used rather than the NCCLS recommended HTM broth. On the basis of current NCCLS breakpoints (NCCLS, 1998), the overall percentages of isolates found to be resistant or intermediate to the following
a Indicates statistically significant difference (P ⬍ 0.05 by chi-square test) in prevalence of -lactamase-positive or BLNAR isolates. b CSF: cerebrospinal fluid.
colony forming U/ml). The trays were incubated at 35°C in ambient air for 18 to 24 h before determining the results. Inoculum size was confirmed by performing viable organism counts. Susceptibility testing medium broth was prepared by using cation-adjusted Mueller–Hinton broth (Becton–Dickinson, MD, USA) with 3% lysed horse blood, 15 mg of NAD per ml, and 5 mg of yeast extract per ml (Eiken, Tokyo, Japan). Antibiotic panels were obtained commerTable 2 MICs of 10 antimicrobials for recent clinical isolates of H. influenzae Antimicrobial agent
Ampicillin Amoxicillin-clavulanatea Cefaclor Cefixime Cefotaxime Ceftriaxone Imipenem Norfloxacin Erythromycin Chloramphenicol a
Total no. of isolates
867 867 732 732 867 867 732 867 867 867
2 : 1 ratio of amoxicillin to clavulanate.
MIC (g/ml)
-Lactamase-negative isolates
-Lactamase-positive isolates
No. of isolates
50%
90%
Range
No. of isolates
50%
90%
Range
756 756 638 638 756 756 638 756 756 756
0.25 0.75 2 0.06 0.03 ⱕ0.015 1 ⱕ0.06 4 0.5
1 3 ⬎16 0.13 0.25 0.03 2 0.13 8 1
ⱕ0.13⬃16 ⱕ0.19⬃24 ⱕ0.25⬃⬎16 ⱕ0.03⬃4 ⱕ0.015⬃4 ⱕ0.015⬃1 ⱕ0.06⬃8 ⱕ0.06⬃1 ⱕ0.13⬃⬎16 ⱕ0.13⬃8
111 111 94 94 111 111 94 111 111 111
⬎16 1.5 8 0.06 0.03 ⱕ0.015 1 ⱕ0.06 4 1
⬎16 6 ⬎16 0.25 0.25 0.03 4 0.13 8 16
4⬃⬎16 0.38⬃12 1⬃⬎16 ⱕ0.03⬃⬎4 ⱕ0.015⬃0.25 ⱕ0.015⬃0.25 0.13⬃8 ⱕ0.06⬃4 2⬃16 0.25⬃16
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
agents were: 13.0% to cefaclor, 1.5% to cefixime, 0.3% to cefotaxime, 0% to ceftriaxone, 1.8% to imipenem, and 4.2% to chloramphenicol. The MICs of erythromycin at which 50% of isolates are inhibited (MIC50) and MIC90 were 4 and 8 g/ml, respectively. The MIC50 and MIC90 of norfloxacin were ⱕ0.06 and 0.13 g/ml, respectively. The norfloxacin MICs of one isolate was 4 g/ml. The following chloramphenicol MICs were obtained with -lactamase-negative and -positive strains, respectively: MIC50, 0.5 and 1 g/ml; MIC90, 1 and 16 g/ml; and the number of strains and the percentage of MICs ⱖ8 g/ml, 1 (0.1%) and 33 (29.7%). The activity of chloramphenicol against -lactamase-positive isolates was significantly diminished compared with that against -lactamase-negative isolates. The activity of cefaclor against -lactamase-positive isolates was slightly diminished compared with that against -lactamase-negative isolates. On the basis of a comparison of the MIC50 and MIC90, -lactamase production did not influence the activity of cefixime, cefotaxime, ceftriaxone, or norfloxacin. A total of 26 -lactamase-negative isolates (3.4% of all -lactamase-negative isolates and 3.0% of all isolates) was found to be BLNAR. The actual distribution of MICs for BLNAR isolates were 4 g/ml, 18; 8 g/ml, 6; and 16 g/ml, two strains. Isolation of BLNAR with high-level resistance (ampicillin-MICs ⱖ 8 g/ml) has also been gradually increasing (1997, 0; 1998, 3; 1999 [until July], five strains). The number of -lactamase-producing isolates for which the ampicillin-MICs were 4, 8, 16, and ⬎16 g/ml were four (3.6%), nine (8.1%), six (5.4%), and 92 strains (82.9%), respectively. Eight -lactamase-positive isolates (7.2% of all -lactamase-producers) were BLPACR. For all eight BLPACR strains, ACV-MICs were 8/4 g/ml, and ampicillin-MICs were ⱖ64 g/ml. Determination of ampicillin- and ACV-MICs for BLNAR or BLPACR were repeated twice, and was never varied by more than ⫾ 1 twofold concentration (data not shown). The -lactamase production was also confirmed by repeat testing. Thus, the overall resistance to ampicillin was 15.8%, and that to ACV was 3.0%. None of the 26 BLNAR and 8 BLPACR isolate were serologically typeable strains, and most of them were recovered from the respiratory tract. The distribution of ACV-MICs for ampicillin-nonsusceptible isolates (ampicillin-MICs ⱖ 2 g/ml; n ⫽ 43) were 2/1 g/ml, 10; 4/2 g/ml, 15; 8/4 g/ml, 13; and 16/8 g/ml, five strains, whereas the modal ampicillin- and ACV-MICs for -lactamase-negative ampicillin-susceptible isolates (ampicillin-MICs ⱕ 1 g/ml) were 0.25 and 0.5/ 0.25 g/ml, respectively. The modal MICs of cefaclor, cefixime, cefotaxime, ceftriaxone, and imipenem for ampicillin-nonsusceptible isolates were ⬎16, 0.06 to 0.13, 0.25, 0.03, and 4 to 8 g/ml, respectively. By comparison, the modal MICs obtained with the same agents for ampicillinsusceptible isolates were two to four, ⱕ0.03, 0.03, ⱕ0.015, and 0.5 to 1 g/ml, respectively. Figs. 1a and b illustrate the distribution of antimicrobial susceptibilities of H. influenzae by -lactamase production
251
and ampicillin-MICs levels. The sensitivity distribution showed a trend toward higher MICs of cephalosporins for -lactamase-negative isolates and correspondingly the rate of isolates with higher ampicillin-MICs also increased. The MICs distribution of cephems for ampicillin-nonsusceptible isolates shifted from 2 to 6 dilutions higher than those for other -lactamase-negative isolates. Fig. 2 shows comparison of cumulative MICs of four cephems for ampicillinnonsusceptible isolates and ampicillin-susceptible strains. Based on the comparison of MICs distribution, ceftriaxone was the most active against H. influenzae in the cephems examined in this study. MICs of ceftriaxone tended to be from 2- to fourfold lower than that of cefotaxime for both ampicillin-nonsusceptible isolates and -lactamase producers. However, both cefotaxime- and ceftriaxone-MICs for ampicillin-nonsusceptible isolates shifted toward higher MICs just as the other cephems did.
4. Discussion H. influenzae causes a variety of community-acquired bacterial infections, especially in children. Nontypeable strains of H. influenzae are the most frequent pathogens causing bronchopneumonia, bronchitis, sinusitis, and acute otitis media. In Japan, H. influenzae serotype b (Hib) still remains the most significant cause of bacterial meningitis, septic arthritis, and epiglottitis in children, because Hib vaccines are not currently available (Uehara et al., 1998). The most common mechanism of ampicillin-resistance with H. influenzae is the production either of two -lactamase, TEM-1 or ROB-1 (Scriver et al., 1994). The results of this study indicated that the overall prevalence of -lactamase production among clinical isolates of H. influenzae was much lower than that reported in the US study (12.8% versus 36.0% for 1994 to 1995) (Jones et al., 1997). Recently, in the SENTRY antimicrobial surveillance program of 837 respiratory tract isolates, the percentage of -lactamase-producing H. influenzae was 33.5% in North America (Doern et al., 1999). In Japan, after comparing an earlier study (Nakamura et al., 1995), it seems that the prevalence of -lactamase-mediated ampicillin-resistance of H. influenzae is not increasing among children. The prevalence of BLPACR’s was similar to that reported in the US (0.8% versus 1.1, 0.2%) (Doern et al., 1997, 1999). However, the prevalence of BLNAR’s found in this study was somewhat higher than that reported in the US (3.0% versus 0.8, 0.1%). In addition, the finding in these surveillance studies have not confirmed the increasing incidents of BLNAR’s in the US (0.8% in 1994 –1995, 0.1% in 1997) (Doern et al., 1997, 1999). Of special interest to us is the prevalence of BLNAR’s increased remarkably to 8.9% during the last 7-month period. The ampicillin-MICs of BLNAR’s have also been gradually increasing. Furthermore, the in vitro activities were significantly diminished against these
252
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
Fig. 1. Frequency distribution of antimicrobial MICs values for H. influenzae isolates, by -lactamase production and ampicillin MICs-levels.
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
253
Fig. 2. Comparison of cumulative MICs four cephems for ampicillin-nonsusceptible isolates (Solid line; ampicillin MICs ⱖ 2 g/ml) and ampicillinsusceptible isolates (Dotted line; ampicillin MICs ⱕ 1 /ml) of H. influenzae.
strains, not only ACV, the -lactamase inhibitor combination, but also other penicillins and cephems. These BLNAR isolates seem to have altered penicillinbinding proteins (PBPs) with decreased affinity or diminished permeability for ampicillin and other -lactams. The characteristics of phenotypic expressions that were observed in this study were consistent with this mechanism. Claioux et al. (1992) reported that changes in PBPs-3A and -3B are a common mechanism that results in a significant level of non-enzymatic resistance to -lactams in H. influenzae regardless of serotype, origin of the isolates, or geographic distribution. Moreover, the level of resistance is related to the extent of PBP modifications. On the other hand, several studies have reported a strain with PBPs-4 and -5 as having a decreased affinity for -lactams (Mendelman et al., 1984; Serfass et al., 1986; Mendelman et al., 1990). On the basis of these previous studies, for a more reliable and effective interpretation of BLNAR, the genes responsible for PBP modification should be clarified as the same as demonstrated in Streptococcus pneumoniae such as pbp2x, pbp1a, and pbp2b (Ubukata et al., 1996; Asahi et al., 1998) The results of this study corroborate the finding in the aforementioned investigation (Doern et al., 1999) regarding the rate of resistant to cefaclor (13.0% versus 12.8%). However, it is interesting that chloramphenicol-resistance of our study was much higher than that reported in the US (4.2% versus 0.7%). Two prior multicenter studies in the US (Doern et al., 1997; Jorgensen et al., 1990) reported chloramphenicol-resistant strains were also uncommon (0.2% and 0%, respectively). With the notable observations in this study, the majority (92%) of the intermediate or resistant isolates for chloramphenicol was -lactamase producers. Thus, chloramphenicol-resistance was more common among -lactamase-positive isolates than -lactamase-negative isolates in Japan (29.7% versus 0.4%). The fluoroquinolones exhibited good activity against both BLNAR’s and BLPACR’s. However, as observed in
this study, fluoroquinolone-resistant H. influenzae strains have been reported by several investigators (Georgiou et al., 1996; Bootsma et al., 1997; Vila et al., 1999). According to these studies, the mechanisms of fluoroquinolone-resistance are mainly due to chromosomal mutations in the gyrA and parC genes, which encode the A subunits of the DNA gyrase and topoisomerase IV, respectively. It will be required to examine the MICs of fluoroquinolone or these gene mutations for isolates recovered from patients with recurrent H. influenzae infections after or during fluoroquinolone therapy. It will be also important to monitor the susceptibility of H. influenzae as their frequent use. Indeed, the actual clinical implications of elevated -lactam MICs for BLNAR or BLPACR strains is poorly understood. One case report has indicated that BLNAR was recovered from the cerebrospinal fluid and its altered PBPs for -lactam were associated with cefuroxime treatment failure (Mendelman et al., 1990). If the frequency of BLNAR’s continue to increase, the incidence of clinical treatment failure cases would be increase in the near future. In our study, none of the BLNAR’s or BLPACR’s was serologically typeable strains and almost all of them were from respiratory tract infections. If BLNAR’s or BLPACR’s emerge among Hib isolates, clinical problems would become much more critical, because Hib vaccines are not currently used in Japan. In conclusion, it is important for laboratory strategies to examine routinely not only -lactamase production, but also to carry out in vitro susceptibility testing, especially MIC determination. In addition, continued monitoring of susceptibility trends will be required to guide appropriate antimicrobial chemotherapy. References Asahi, Y. & Ubukata, K. (1998). Association of a Thr-371 substitution in a conserved amino acid motif of penicillin-binding protein 1A with
254
K. Ohkusu et al. / Diagnostic Microbiology and Infectious Disease 36 (2000) 249 –254
penicillin resistance of Streptococcus pneumoniae. Antimicrob Agents Chemother 42, 2267–2273. Barry, A. L., Fuchs, P. C., & Pfaller, M. A. (1993). Susceptibilities of -lactamase-producing and -nonproducing ampicillin-resistant strains of Haemophilus influenzae toceftibuten, cefaclor, cefuroxime, cefixime, cefotaxime, and amoxicillin-clavulanic acid. Antimicrob Agents Chemother 37, 14 –18. Bootsma, H. J., Troeltra, A., Veen–Rutgers, A., Mooi, F. R., Neeling, A. J., & Overbeek, B. P. (1997). Isolation and characterization of a ciprofloxacin-resistant isolate of Haemophilus influenzae from The Netherlands. J Antimicrob Chemother 39, 292–293. Campos, J. M. (1995). Haemophilus. In P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, & R. H. Yolken (Eds.). Manual of Clinical Microbiology, 6th edition (pp. 556 –565). Washington, D.C.: ASM Press. Clairoux, N., Picard, M., Brochu, A., Rousseau, N., Gourde, P., Beauchamp, D., Parr, T. R., Bergeron, M. G., & Malouin, F. (1992). Molecular basis of the non- -lactamase-mediated resistance to -lactam antibiotics in strains of Haemophilus influenzae isolated in Canada. Antimicrob Agents Chemother 36, 1504 –1513. Doern, G. V., Brueggemann, A. B., Pierce, G., Holley, H. P., & Rauch, A. (1997). Antibiotic resistance among clinical isotates of Haemophilus influenzae in the United States in 1994 and 1995 and detection of -lactamase-positive strains resistant to amoxicillin-clavulanate: results of a national multicenter surveillance study. Antimicrob Agents Chemother 41, 292–297. Doern, G. V., Jones, R. N., Pfaller, M. A., Kugler, K., & the SENTRY participants group (1999). Haemophilus influenzae and Moraxella catarrhalis from patients with community-acquired respiratory tract infections: antimicrobial susceptibility patterns from the SENTRY antimicrobial surveillance program (US and Canada, 1997). Antimicrob Agents Chemother 43, 385–389. Georgiou, M., Munoz, R., Roman, F., Canton, R., Gomez–Lus, R., Campos, J., & Campa, A. G. (1996). Ciprofloxacin-resistant Haemophilus influenzae strains possess mutations in analogous positions of gyrA and. parC Antimicrob Agents Chemother 40, 1741–1744. Ishiwada, N., Kuroki, H., Ishikawa, N., Sugimoto, K., Koori, Y., Suruga, Y., Uehara, S., & Niimi, H. (1998). Characteristics of -lactamaseproducing and amoxicillin-clavulanate-resistant strains of Haemophilus influenzae isolated from pediatric patients. J Infect Chemother 4, 112– 115. Jones, R. N., Jacobs, M. R., Washington, J. A., & Pfaller, M. A. (1997). A 1994 –95 survey of Haemophilus influenzae susceptibility to ten orally
administered agents A 187 clinical laboratory center sample in the United States. Diagn Microbiol Infect Dis 27, 75– 83. Jorgensen, J. H., Doern, G. V., Maher, L. A., Howell, A. W., & Redding, J. S. (1990). Antimicrobial resistance among respiratory isolates of Haemophilus influenzae, Moraxella catarrhalis, and Streptococcus pneumoniae in the US. Antimicrob Agents Chemother 34, 2075–2080. Mendelman, P. M., Chaffin, D. O., Stull, T. L., Rubens, C. E., Mack, K. D., & Smith, A. L. (1984). Characterization of non--lactamase-mediated ampicillin resistance in Haemophilus influenzae. Antimicrob Agents Chemother 26, 235–244. Mendelman, P. M., Chaffin, D. O., Krilov, L. R., Kalaitzoglou, G., Serfass, D. A., Onay, O., Wiley, E. A., Overturf, G. D., & Rubin, L. G. (1990). Cefuroxime treatment failure of nontypable Haemophilus influenzae meningitis associated with alteration of penicillin-binding proteins. J Infect Dis 162, 1118 –1123. Nakamura, A., Ishikawa, N., Suzuki, H., Kurosaki, T., Terashima, I., Uehara, S., & Niimi, H. (1995). Clinical bacteriology of Haemophilus influenzae isolated from respiratory tract infections in children. J Jpn Pediatr Soc 99, 1234 –1241. National Committee for Clinical Laboratory Standards (NCCLS) (1998). Performance standards for antimicrobial susceptibility testing. Supplement tables, M100-S8. Wayne, PA: NCCLS. Scriver, S. R., Walmsley, S. L., Kau, C. L., Hoban, D. J., Brunton, J., McGeer, A., Moore, T. C., Witwicki, E., Canadian Haemophilus Study Groop, & Low, D. E. (1994). Determination of antimicrobial susceptibilities of Canadian isolates of Haemophilus influenzae and characterization of their -lactamases. Antimicrob Agents Chemother 38, 1678 –1680. Serfass, D. A., Mendelman, P. M., Chaffin, D. O., & Needman, C. A. (1986). Ampicillin resistance and penicillin-binding proteins of Haemophilus influenzae. J Gen Microbiol 132, 2855–2861. Ubukata, K., Asahi, Y., Yamane, A., & Konno, M. (1996). Combinational detection of autolysin and penicillin-binding protein 2B genes of Streptococcus pneumoniae by PCR. J Clin Microbiol 34, 592–596. Uehara, S., Kamiya, H., Togashi, T., Kato, T., Shiraki, K., & Morishima, T. (1998). Epidemiology of Haemophilus influenzae meningitis among infants and children in Japan comparison with incidences of bacterial meningitis. J Jpn Pediatr Soc 102, 656 – 665. Vila, J., Ruiz, J., Sanchez, F., Navarro, F., Mirelis, B., Anta, M. E. J., & Prats, G. (1999). Increase in quinolone resistance in a Haemophilus influenzae strain isolated from a patient with recurrent respiratory infections treated with ofloxacin. Antimicrob Agents Chemother 43, 161–162.