Volume 13, Issue 5 , Pages 584-588, September 2009
High prevalence of β-lactam-resistant Haemophilus influenzae type b isolates derived from respiratory tract specimens in Japanese patients
Article Outline
Summary
Objective
Serotypeable strains of Haemophilus influenzae, which can cause invasive infections, are found in the respiratory tract at low frequencies. We compared the antibiotic resistance of the typeable and nontypeable strains of H. influenzae in respiratory tract specimens obtained in Japan.
Methods
We determined the serotypes and the antibiotic susceptibilities of 440 clinical H. influenzae strains isolated from respiratory tract specimens. We also examined the prevalence of genotypes that are associated with β-lactam resistance.
Results
The majority of the strains were nontypeable (421 strains, 95.7%). The remainder belonged to serotypes b (10 strains, 2.3%), e (three strains, 0.7%), or f (six strains, 1.4%). The type b strains exhibited the expression of β-lactamase and resistance mutations in penicillin-binding protein 3 with significantly higher frequencies than other strains.
Conclusions
H. influenzae type b strains, which are associated with meningitis and bacteremia, derived from respiratory tract specimens, shared more β-lactam-resistant mechanisms than nontypeable and other serotype strains.
Keywords: Haemophilus influenzae, Serotype, BLNAR, β-Lactamase, Antibiotic resistance, Respiratory tract specimen
Introduction
Along with Streptococcus pneumoniae, Haemophilus influenzae is a major causative factor of respiratory and otolaryngology infections, especially community-acquired pneumonia in elderly persons and otitis media and sinusitis in children.1, 2 Penicillins and second- or third-generation cephalosporins (CEPs) are used as the first-choice antibiotics for these infectious diseases. There has been a recent worldwide increase in infections with two types of β-lactam-resistant H. influenzae strains. One is due to the expression of β-lactamases, and another is mutations in penicillin-binding proteins (PBPs), especially PBP3.3, 4 With regard to the β-lactamase-carrying H. influenzae strains, two types of β-lactamases denoted as TEM-1 and ROB-1 have been observed. TEM-1 is highly dominant in Japan compared to ROB-1.5 With regard to the strains that only have mutations in PBPs and no β-lactamases – these are called β-lactamase-negative ampicillin-resistant (BLNAR) strains – there has been a recent and highly problematic increase in community-acquired infections with these strains.4, 6 BLNAR strains are highly prevalent in Japan.
Most H. influenzae clinical isolates from respiratory tract infections do not carry capsular polysaccharides and are thus nontypeable. However, the typeable strains, which are less prevalent, are associated with invasive infectious diseases such as meningitis and bacteremia. This is because they bear capsular polysaccharides that facilitate their escape from complement-mediated killing and phagocytosis.7 Of the typeable strains, type b strains occur most frequently and are responsible for more than half of all cases of pediatric bacterial meningitis in Japan.8 H. influenzae type b (Hib) vaccine was introduced in Japan in 2007, so most children have never been vaccinated.
Our ability to treat invasive H. influenzae infections with antibiotics will be threatened by the increasing prevalence of antibiotic-resistant typeable H. influenzae strains. This study was undertaken in order to determine the extent of this problem; we examined susceptibility to various antibiotics and the β-lactam resistance genotypes of clinical H. influenzae isolates obtained from respiratory tract specimens provided by clinical laboratories of general hospitals and a commercial clinical laboratory in Japan.
Materials and methods
Bacterial strains
Between 2002 and 2004, 440 H. influenzae clinical strains were isolated from respiratory tract specimens obtained in hospitals in the Hokkaido Prefecture, Japan. The strains were collected and stocked by Sapporo Clinical Laboratory, Inc. (Sapporo, Japan), Muroran General Hospital (Muroran, Japan), and Hokkaido University Hospital (Sapporo, Japan). This study was approved by the respective review boards of the institutions and the corporation. Sapporo Clinical Laboratory, Inc. serves almost the entire area of the Hokkaido Prefecture. Due to the Act on the Protection of Personal Information in Japanese law, we were only given the following information: patient age and sex, the clinical source, and the name of the city in which the strains were isolated. The isolates obtained from the respiratory specimens (sputa, rhinorrhea, pharyngeal fluids, nasal cavity, otopyorrhea, and bronchial lavage fluids) were examined in this study. The strains were identified as H. influenzae with WalkAway40 (Dade Behring, Tokyo, Japan). We also tested X and Y growth requirements by a disk method (Eiken Chemical, Tokyo, Japan). All isolates were grown at 37
°C in an atmosphere of 5% CO2 on a chocolate agar II plate (Nippon Becton-Dickinson, Tokyo, Japan).
Analysis of genotypes contributing to β-lactam resistance
Genomic DNA isolated from the cells grown on chocolate agar served as the template for PCR analysis that was performed according to Hasegawa et al.9 PCR amplification was performed by using HotStarTaq DNA polymerase (Qiagen, Hilden, Germany). The PCR detected the following genes: outer membrane protein P6 (as a positive control for H. influenzae), TEM-1 β-lactamase, and ROB-1 β-lactamase. The genotypes of β-lactamase-negative ampicillin low resistant (gLow-BLNAR) and β-lactamase-positive amoxicillin/clavulanic acid low resistant (gLow-BLPACR) strains that bear mutations in PBP3 that confer low-level β-lactam resistance were detected by identifying the Asn526Lys amino acid substitution. The genotypes of β-lactamase-negative ampicillin high resistant (gHigh-BLNAR) and β-lactamase-positive amoxicillin/clavulanic acid high resistant (gHigh-BLPACR) strains that bear mutations in PBP3 that confer high-level β-lactam resistance were detected by identifying the Ser385Thr and Asn526Lys amino acid substitutions. The strains that did not contain the PBP3 resistance mutations were genotypes β-lactamase-negative ampicillin susceptible (gBLNAS) and β-lactamase-positive amoxicillin/clavulanic acid susceptible (gBLPACS).
Serotype determination
Serotype was determined by the slide agglutination assay using specific antisera (Denka Seiken, Tokyo, Japan) and by PCR performed according to the method of Falla et al.10
β-Lactamase assay (disk method)
Strains were tested for β-lactamase production by using nitrocefin-impregnated disks (Becton Dickinson, Cockeysville, MD, USA).
Antibiotic susceptibility
Determination of the minimum inhibitory concentration (MIC) of antibiotics according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI, Wayne, PA, USA)11 was performed as described previously.12
Statistical analysis
The data were statistically analyzed using the Fisher's exact test.
Results
A total of 440 H. influenzae strains were isolated from respiratory tract specimens obtained between 2002 and 2004 from patients in the Hokkaido Prefecture, Japan. Of these strains, 421 were nontypeable (95.7%), and 10 (2.3%), three (0.7%), and six (1.4%) belonged to serotypes b, e, and f, respectively. The patients were mainly pediatric (64.5%) and the elderly (21.6%). The distribution of patient age was not significantly different among serotypes (Table 1). Furthermore, the serotype distribution did not differ significantly in terms of clinical source (Table 2) and city from which the strains were isolated (data not shown). The expression of β-lactamase by the 440 strains was determined by PCR and the disk method. The results obtained from both methods were consistent (data not shown). All β-lactamases detected were of the TEM-1 type. Although only 5.5% of the nontypeable strains expressed β-lactamase, 50% of the type b strains expressed it (Table 3). This difference was statistically significant (p<0.01). β-Lactamase was not detected in any of the serotype e or f strains.
Table 1. Relationship between patient age and Haemophilus influenzae serotype.
| Age (years) | Number of strains | ||||
|---|---|---|---|---|---|
| Nontypeable | b | e | f | Total | |
| <16 | 270 (64.1%) | 7 (70%) | 2 (66.7%) | 5 (83.3%) | 284 (64.5%) |
| 16–59 | 61 (14.5%) | 0 | 0 | 0 | 61 (13.9%) |
| >59 | 90 (21.4%) | 3 (30%) | 1 (33.3%) | 1 (16.7%) | 95 (21.6%) |
| Total | 421 (100%) | 10 (100%) | 3 (100%) | 6 (100%) | 440 (100%) |
Table 2. Relationship between type of clinical specimen and Haemophilus influenzae serotype.
| Specimen | Number of strains | ||||
|---|---|---|---|---|---|
| Nontypeable | b | e | f | Total | |
| Sputum | 183 | 5 | 2 | 0 | 190 (43.2%) |
| Rhinorrhea | 170 | 3 | 1 | 0 | 174 (39.5%) |
| Pharyngeal fluid | 32 | 0 | 0 | 1 | 33 (7.5%) |
| Nasal cavity | 19 | 0 | 0 | 0 | 19 (4.3%) |
| Otopyorrhea | 12 | 2 | 0 | 4 | 18 (4.1%) |
| Bronchial lavage fluid | 5 | 0 | 0 | 1 | 6 (1.4%) |
| Total | 421 | 10 | 3 | 6 | 440 (100%) |
Table 3. Relationship between the expression of β-lactamase and Haemophilus influenzae serotype.
| β-Lactamase | Number of strains | ||||
|---|---|---|---|---|---|
| Nontypeable | b | e | f | Total | |
| Positive | 23 (5.5%) | 5a (50%) | 0 (0%) | 0 (0%) | 28 (6.4%) |
| Negative | 398 (94.5%) | 5 (50%) | 3 (100%) | 6 (100%) | 412 (93.6%) |
| Total | 421 (100%) | 10 (100%) | 3 (100%) | 6 (100%) | 440 (100%) |
ap |
With regard to the resistance mutations in PBP3 (Table 4), 10% of type b and 50.8% of nontypeable strains had PBP3 without resistance mutations (gBLNAS and gBLPACS strains). This difference was statistically significant (p<0.05). The PBP3 mutation that results in low-grade resistance (found in gLow-BLNAR and gLow-BLPACR strains) was detected in 60% of type b and 20.4% of nontypeable strains. This difference was also statistically significant (p<0.01). The type b strains did not differ from the nontypeable strains in terms of the frequency of PBP3 mutations that lead to high-grade resistance (found in gHigh-BLNAR and gHigh-BLPACR strains; 30% vs. 28.7%). Thus, the low-level resistance mutation in PBP3 was more frequently found in the type b strains than the nontypeable strains.
Table 4. Relationship between penicillin-binding protein 3 (PBP3) genotypes and Haemophilus influenzae serotype.
| Resistance mutations in PBP3 | Number of strains | ||||
|---|---|---|---|---|---|
| Nontypeable | b | e | f | Total | |
| gBLNAS and gBLPACS | 214 (50.8%) | 1a (10%) | 2 (66.7%) | 4 (66.7%) | 221 (50.2%) |
| gLow-BLNAR and gLow-BLPACR | 86 (20.4%) | 6b (60%) | 0 (0%) | 1 (16.7%) | 93 (21.1%) |
| gHigh-BLNAR and gHigh-BLPACR | 121 (28.7%) | 3 (30%) | 1 (33.3%) | 1 (16.7%) | 126 (28.6%) |
| Total | 421 (100%) | 10 (100%) | 3 (100%) | 6 (100%) | 440 (100%) |
ap |
bp |
Table 5 shows the susceptibility of the H. influenzae strains to ampicillin (ABPC). As expected from the data of genotypes described above, the type b strains were significantly more frequently resistant to ABPC (p<0.01). In terms of other antibiotics, all five β-lactamase-positive type b strains were resistant to tetracycline (≥8μg/ml), and one of these strains was intermediate to chloramphenicol (4μg/ml). All type b strains were susceptible to clarithromycin (MIC 4–8μg/ml), levofloxacin (≤0.12μg/ml), meropenem (≤0.12μg/ml), and rifampin (≤0.5μg/ml). With regard to nontypeable strains, eight (1.9%) tetracycline, three (0.2%) chloramphenicol, eight (1.9%) levofloxacin, one (0.2%) meropenem, and no rifampin-resistant strains were observed. No resistant strains of these antibiotics were observed in type e or f strains.
Table 5. Relationship between ampicillin (ABPC) susceptibility and Haemophilus influenzae serotype.
| Susceptibility to ABPC | Number of strains | ||||
|---|---|---|---|---|---|
| Nontypeable | b | e | f | Total | |
| Susceptible | 291 (69.1%) | 3 (30%) | 2 (66.7%) | 4 (66.7%) | 300 (68.2%) |
| Intermediate | 78 (18.5%) | 1 (10%) | 1 (33.3%) | 2 (33.3%) | 82 (18.6%) |
| Resistant | 52 (12.4%) | 6a (60%) | 0 (0%) | 0 (0%) | 58 (13.2%) |
| Total | 421 (100%) | 10 (100%) | 3 (100%) | 6 (100%) | 440 (100%) |
ap |
Discussion
In the present study, we examined 440 H. influenzae clinical isolates obtained from respiratory tract specimens from Japanese patients. The vast majority (95.7%) of the strains were nontypeable. The remainder belonged to serotypes b (2.3%), f (0.7%), and e (1.4%). The prevalence of the serotypeable strains observed here is comparable to those estimated in other Japanese studies.13, 14 We found that 50% of the serotype b strains expressed β-lactamase and 90% had some resistance mutations in PBP3, and that these frequencies were significantly higher than those in the nontypeable strains (5.5% and 49.1%, respectively) and the strains with other serotypes (0% and approximately 33%, respectively).
The typeable strains can cause invasive infections such as meningitis and bacteremia because they evade complement-dependent bacterial killing and phagocytosis by a function of their capsular polysaccharide.7 In contrast, the nontypeable strains, which do not bear capsular polysaccharides, are rapidly killed in the blood in a complement-dependent manner. The type b strains are still the most frequently isolated of the six serotype strains in Japan, because H. influenzae type b vaccine has only just been introduced on 2007. Recent nationwide surveillance for bacterial meningitis in Japan (2000–2004) reported that gBLNAR increased in recent years, and approximately 70% H. influenzae strains shared resistance mutations in PBP3.15, 16 Approximately 20% type b strains isolated from meningitis shared β-lactamase; however, β-lactamase-shared strains tended to decrease every year in Japan (they decreased from about 25% in 2000–2002 to 15% in 2004).16 Nevertheless, the positive β-lactamase rate of H. influenzae was higher in strains from meningitis than in those from other diseases. The β-lactamase-positive strains derived from respiratory diseases (mainly nontypeable strains) had a prevalence of 15–20% until 1995, after which the prevalence gradually declined to 3% in 1999.6 The reason for this decrease is considered to be the predominant use of oral third-generation CEPs, which are stable with respect to β-lactamase.6
With regard to H. influenzae type b strains derived from the respiratory tract, Nariai reported a prevalence of 56.7% of BLNAR in type b strains and low frequency (6.7%) occurrence of β-lactamase derived from the nasopharynx of children.17 In contrast, our study indicated that strains that were derived from respiratory specimens shared β-lactamase with a significantly higher rate (50%) compared to nontypeable and other typeable strains. A similar observation was reported by Luong et al.14 They reported that all three type b strains tested expressed β-lactamase; however, the sample numbers in that study were relatively low. Our present study also showed that mutations in PBP3 are found to have a higher prevalence in low-grade resistance mutations, namely gLow-BLNAR and gLow-BLPACR, in type b strains compared to other strains. The reason for the differences among these reports is not yet clear. All β-lactamase-positive type b strains in the present study also shared tetracycline resistance. So a resistant type b clone or a plasmid carrying both β-lactamase and tetracycline-resistant genes4 may spread. On the other hand, the use of oral third-generation CEPs could cause a decrease in β-lactamase-positive strains and an increase in BLNAR. Further surveys on antibiotic usage are required to clarify this. The increase in High-BLNAR in meningitis is of concern.16 In conclusion, we observed a high frequency of resistance in H. influenzae type b strains isolated from respiratory tract specimens in a prefecture of Japan. It is feared that such antibiotic-resistant type b strains, which can cause invasive infections such as meningitis, are widespread.
Acknowledgements
We thank Tasuku Hayashi, Keiko Matsuda (Muroran General Hospital), Osamu Kuwahara (Sapporo Clinical Laboratory), and Hirotsugu Akizawa (Hokkaido University Hospital) for the H. influenzae clinical isolates. This work was partially supported by a grant from the Yuasa Memorial Foundation.
Conflict of interest: No conflict of interest to declare.
References
- Prospective multicenter study of the causative organisms of community-acquired pneumonia in adults in Japan. J Infect Chemother. 2006;12:63–69
- . Current status of bacterial resistance in the otolaryngology field: results from the Second Nationwide Survey in Japan. J Infect Chemother. 2003;9:46–52
- Association of amino acid substitutions in penicillin-binding protein 3 with β-lactam resistance in β-lactamase-negative ampicillin-resistant Haemophilus influenzae. Antimicrob Agents Chemother. 2001;45:1693–1699
- . Antimicrobial resistance in Haemophilus influenzae. Clin Microbiol Rev. 2007;20:368–389
- . Molecular evolution of β-lactam-resistant Haemophilus influenzae: 9-year surveillance of penicillin-binding protein 3 mutations in isolates from Japan. Antimicrob Agents Chemother. 2006;50:2487–2492
- . Problems associated with high prevalence of multidrug-resistant bacteria in patients with community-acquired infections. J Infect Chemother. 2003;9:285–291
- . Participation of complement in host defense against capsule-deficient Haemophilus influenzae. Infect Immun. 1983;42:708–715
- Trends in pediatric bacterial meningitis in Japan (2003–2004) [In Japanese]. Kansenshogaku Zasshi. 2006;80:27–38
- Diversity of ampicillin-resistance genes in Haemophilus influenzae in Japan and the United States. Microb Drug Resist. 2003;9:39–46
- . PCR for capsular typing of Haemophilus influenzae. J Clin Microbiol. 1994;32:2382–2386
- Clinical and Laboratory Standards Institute. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. M07-A7 approved standard. 7th ed. Wayne, PA: CLSI; 2006.
- Antibiotic susceptibility of Haemophilus influenzae strains isolated from various clinical sources in Hokkaido Prefecture, Japan. J Infect Chemother. 2008;14:93–98
- . Surveillance based on molecular epidemiology for Streptococcus pneumoniae isolates between 1998 and 2000 in Japan—results of clinical isolates collected by the Community-Acquired Bacterial Infections Working Group [In Japanese]. Jpn J Chemother. 2003;51:60–70
- . Serotypes of Haemophilus influenzae strains isolated from pediatric patients with respiratory tract infections. Tohoku J Exp Med. 2004;202:245–254
- Rapidly increasing prevalence of β-lactamase-nonproducing, ampicillin-resistant Haemophilus influenzae type b in patients with meningitis. Antimicrob Agents Chemother. 2004;48:1509–1514
- High prevalence of type b β-lactamase-non-producing ampicillin-resistant Haemophilus influenzae in meningitis: the situation in Japan where Hib vaccine has not been introduced. J Antimicrob Chemother. 2006;57:1077–1082
- . Prevalence of β-lactamase-nonproducing ampicillin-resistant Haemophilus influenzae and Haemophilus influenzae type b strains obtained from children with lower respiratory tract infections. J Infect Chemother. 2007;13:396–399
PII: S1201-9712(08)01558-0
doi:10.1016/j.ijid.2008.09.017
© 2008 International Society for Infectious Diseases. Published by Elsevier Inc. All rights reserved.
Volume 13, Issue 5 , Pages 584-588, September 2009
