- Email: [email protected]
Association between PCR ribotypes and antimicrobial susceptibility among Clostridium difficile isolates from healthcare-associated infections in South Korea
International Journal of Antimicrobial Agents 40 (2012) 24–29
Contents lists available at SciVerse ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Association between PCR ribotypes and antimicrobial susceptibility among Clostridium difficile isolates from healthcare-associated infections in South Korea Jieun Kim a , Jung Oak Kang b,∗ , Hyunjoo Pai a , Tae Yeal Choi b a b
Department of Internal Medicine, Hanyang University College of Medicine, Seoul, South Korea Department of Laboratory Medicine, Hanyang University College of Medicine, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 2 March 2012 Accepted 22 March 2012 Keywords: Clostridium difficile Antimicrobial susceptibility Ribotyping
a b s t r a c t In this study, the association between antimicrobial susceptibility, PCR ribotype and presence of the ermB gene in clinical isolates of Clostridium difficile was investigated. PCR ribotyping and ermB gene PCR were performed on 131 C. difficile isolates. The susceptibility of these isolates to metronidazole, vancomycin, piperacillin/tazobactam (TZP), clindamycin, moxifloxacin and rifaximin was also determined. Use of antibiotics within the previous 2 months was documented. Resistance rates to clindamycin, moxifloxacin and rifaximin were 67.9%, 62.6% and 19.1%, respectively. No metronidazole, vancomycin or TZP resistance was detected. Previous exposure to moxifloxacin was significantly correlated with resistance to this antibiotic, but prior use of clindamycin was not significantly correlated with clindamycin resistance. Sixty-four strains (48.9%) carried the ermB gene, of which all but one (98.5%) were resistant to clindamycin. The clindamycin resistance rates of the common PCR ribotypes (018, 017 and 001) were 91.4%, 100% and 84.2%, respectively, and their moxifloxacin resistance rates were 91.4%, 95.0% and 78.9%, respectively. Resistance rates to rifaximin were 5.7% and 95.0% in ribotype 018 and 017 strains, whilst none of the 001 strains were resistant to rifaximin. In conclusion, the common ribotypes 018, 017 and 001 of C. difficile have high rates of resistance to clindamycin and moxifloxacin, but differ greatly in the frequency of rifaximin resistance. © 2012 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction
both [4]. The most commonly used antimicrobial in the 1970s was clindamycin, and in the 1980s it was cephalosporins, but recent epidemic cases in Canada suggest that fluoroquinolones may now play an important role [5]. Therefore, surveillance of prior antimicrobial use, antimicrobial resistance patterns in the community and hospitals, and characterisation of the corresponding molecular mechanisms are very important for avoiding treatment failure or complications as well as preventing nosocomial or community spread of epidemic strains. In this study, the association between recent antimicrobial use and antimicrobial resistance of clinical isolates of toxigenic C. difficile in a tertiary care hospital in South Korea was investigated. Antimicrobial resistance in relation to epidemic PCR ribotype patterns and resistance genes was also analysed.
Clostridium difficile has been known for more than 30 years as the primary cause of antibiotic-associated diarrhoea and pseudomembranous colitis. Currently, a new strain known as type BI from restriction endonuclease analysis, North American pulsed-field type 1 (NAP1) from pulsed-field gel electrophoresis and ribotype 027 from PCR ribotyping has become the single most important epidemic strain causing C. difficile infection (CDI) worldwide [1]. Because of the ribotype 027 strain epidemic, the incidence of CDI, the proportion of complicated cases and fatality rates have all increased in North America and Europe [2]. However, the ribotype 027 strain was rarely reported in South Korea until recently, and molecular epidemiology has instead revealed a nationwide epidemic of A− B+ strains of C. difficile starting in 2003 [3]. There are several explanations for the changing epidemiology of CDI, including alterations of antimicrobial use and the emergence of a new strain with increased virulence, antimicrobial resistance or
2.1. Setting and study design
∗ Corresponding author. Present address: Department of Laboratory Medicine, Hanyang University, Guri Hospital, Gyeongchun-ro 153, Guri City, Gyeonggi-do, 471-701, South Korea. Tel.: +82 31 560 2572; fax: +82 31 560 2585. E-mail address: [email protected] (J.O. Kang).
This study was conducted at Hanyang University Hospital, a 900-bed tertiary care facility located in Seoul, South Korea. From September 2008 through January 2010, all C. difficile isolates from patients older than 18 years of age with healthcare-associated C. difficile infections (HA-CDIs) were enrolled. The study was approved
2. Materials and methods
0924-8579/$ – see front matter © 2012 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. http://dx.doi.org/10.1016/j.ijantimicag.2012.03.015
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
by the institutional review board of Hanyang University Hospital (HYUH IRB 2010-R-12). Informed consent was waived by the board. 2.2. Definitions HA-CDI was defined when the C. difficile isolates from stool cultures were positive for toxin genes (tcdA, tcdB, cdtA or cdtB) by multiplex PCR or gave positive results in toxin antigen assay A&B using a commercial kit (VIDAS® C. difficile Toxins A & B; bioMérieux SA, Marcy l’Étoile, France) in patients with diarrhoea that developed ≥72 h after hospitalisation or within 2 months of the last discharge, provided the patient had not resided in a healthcare facility. Previous use of antimicrobial agents was defined when a patient had used the same class of antibiotics for ≥3 days within the last 2 months. Use of clindamycin, fluoroquinolones (moxifloxacin, levofloxacin and ciprofloxacin), rifamycins (rifampicin and rifaximin), metronidazole, glycopeptides (vancomycin and teicoplanin) and -lactam/-lactamase inhibitors [piperacillin/tazobactam (TZP), ampicillin/sulbactam and amoxicillin/clavulanic acid] was investigated. 2.3. Isolation of Clostridium difficile
25
Resistance breakpoints were as defined by the Clinical and Laboratory Standards Institute (CLSI) [8] (Table 1). For those antimicrobial agents with no defined breakpoint, resistance was defined according to a previous report by Huang et al. [9] (Table 1). Highlevel resistance was defined as MICs for clindamycin of ≥256 mg/L and for moxifloxacin of ≥32 mg/L [9]. 2.6. Detection of the ermB gene ermB gene-specific PCR was performed with template DNA, as described previously, with primers 2980 (5 AATTAAGTAAACAGGTAACGTT3 ) and 2981 (5 GCTCCTTGGAAGCTGTCAGTAG3 ) [10]. A PCR product of 688 bp on electrophoresis was considered a positive result for ermB. 2.7. Statistical methods SPSS v.13.0 for Windows (SPSS Inc., Chicago, IL) was used for statistical analysis. Categorical variables were analysed by Pearson’s 2 test or Fisher’s exact test. A P-value of <0.05 by two-tailed test was considered statistically significant. 3. Results
Stool specimens were cultured anaerobically on cycloserine– cefoxitin–taurocholate agar (Oxoid Ltd., Cambridge, UK) supplemented with 7% horse blood after alcohol shock treatment of stool specimens. Colonies of C. difficile were identified with an API® Rapid ID 32A system (bioMérieux SA, Lyon, France). 2.4. PCR ribotyping of Clostridium difficile PCR ribotyping was performed using template DNA as described elsewhere [6]. Amplification products were fractionated by electrophoresis through 8% acrylamide gels (30% acrylamide/bis solution, 29:1; Bio-Rad, Richmond, CA) and were visualised under ultraviolet by staining gels for 30 min with ethidium bromide. Banding patterns were checked visually. Each unique pattern was assigned its own ribotype code and was matched with the PCR ribotypes of the reference strains ribotype 027 and ATCC 43598 strain (ribotype 017) [7]. 2.5. Antimicrobial susceptibility tests Susceptibility tests were conducted on 143 isolates from 188 HA-CDI patients, and minimum inhibitory concentrations (MICs) of six antimicrobial agents (metronidazole, vancomycin, TZP, clindamycin, moxifloxacin and rifaximin) were determined for 131 C. difficile isolates from 131 HA-CDI patients. The MICs of clindamycin, moxifloxacin, metronidazole, vancomycin and TZP were determined by Etest (AB-BIODISK, Solna, Sweden) and those of rifaximin were determined by the agar dilution test. Clostridium difficile ATCC 700057 was used as a quality control strain for the susceptibility tests.
Among 131 isolates, 20 toxin B-only, 100 toxin A and B, and 11 binary toxin as well as toxin A and B-producing isolates were included. 3.1. Minimum inhibitory concentration distribution of Clostridium difficile for various antimicrobial agents The MIC distribution patterns of the 131 C. difficile isolates differed depending on the antimicrobial agent (Fig. 1). The MICs of clindamycin (Fig. 1A), moxifloxacin (Fig. 1B) and rifaximin (Fig. 1C) each formed bimodal distributions clearly separated by their breakpoints. The distributions of the MICs for TZP (Fig. 1D), metronidazole (Fig. 1E) and vancomycin (Fig. 1F) were unimodal with narrow ranges of MICs. None of the isolates were resistant to metronidazole, vancomycin or TZP (Table 1). The rates of resistance to clindamycin and moxifloxacin were 67.9% and 62.6%, respectively, and 80 (61.1%) of the isolates were resistant to both drugs. The rates of high-level resistance to clindamycin (≥256 mg/L) and moxifloxacin (≥32 mg/L) were 56.5% and 29.0%, respectively. Twenty-five isolates (19.1%) were resistant to rifaximin, of which 22 (88.0%) were also resistant to clindamycin and 23 (92.0%) to moxifloxacin, with close correlations between the resistances (P = 0.017 and P < 0.0001, respectively). 3.2. Correlation between antibiotic resistance and previous antibiotic exposure Previous use of clindamycin had a marginal effect on the proportion of strains resistant to clindamycin (P = 0.053). On the other
Table 1 Antimicrobial resistance rates of 131 isolates from healthcare-associated Clostridium difficile infection and previous antibiotic history. Antibiotic
Breakpoint (mg/L)a
Resistance rate [n (%)]
Exposure to same class of antibiotics [n (%)]
P-value
Clindamycin Moxifloxacin Rifaximin Metronidazole Vancomycin Piperacillin/tazobactam
≥8 ≥8 ≥4 ≥32 ≥32 ≥128/4
89 (67.9) 82 (62.6) 25 (19.1) 0 (0.0) 0 (0.0) 0 (0.0)
29 (22.1) 59 (45.0) 2 (1.5) 35 (26.7) 27 (20.6) 41 (31.3)
0.053 0.019
a The breakpoints for most antimicrobial agents were considered according to Clinical and Laboratory Standards Institute guidelines [8]. Resistance to rifaximin and vancomycin was considered according to a previous report by Huang et al. [9].
26
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
Fig. 1. Minimum inhibitory concentration (MIC) distribution of 131 isolates of Clostridium difficile to various antimicrobial agents. (A–C) MICs of clindamycin (A), moxifloxacin (B) and rifaximin (C), separated by vertical grey line indicating the breakpoint of each agent. (D–F) Piperacillin/tazobactam (D), metronidazole (E) and vancomycin (F) showed MICs with a unimodal distribution.
hand, previous use of fluoroquinolones was highly related to the frequency of resistance to moxifloxacin (P = 0.019). In the case of rifaximin, none of the patients had previous exposure to rifaximin and only two had been exposed to rifampicin, and these latter were resistant to rifaximin. Previous use of metronidazole, vancomycin and TZP was not associated with resistance.
Eleven binary toxin-producing strains were identified, including three of ribotype 027. Ribotype 027 was not resistant to any of the six antimicrobial agents. None of the other binary toxin-producing isolates was resistant to any of the six antimicrobial agents, except for one with a MIC of 8 mg/L for rifaximin. 3.4. Presence of the ermB gene
3.3. Antibiotic resistance according to PCR ribotypes of Clostridium difficile A total of 35 distinct PCR ribotypes was observed among the 131 isolates, the most common being ribotype 018 (35; 26.7%), followed by ribotype 017 (20; 15.3%) and ribotype 001 (19; 14.5%). MICs for clindamycin, moxifloxacin and rifaximin among the isolates of seven ribotypes are shown in Table 2. The most common ribotype (018) had high rates of resistance to clindamycin (91.4%) and moxifloxacin (91.4%), but a low rate of resistance to rifaximin (5.7%). All of the clindamycin-resistant 018 isolates were also resistant to moxifloxacin. Ribotypes 001 and 112 also had high rates of resistance to clindamycin (84.2% and 100%) and moxifloxacin (78.9% and 75.0%), and none of the isolates was resistant to rifaximin. All of the 001 and 112 isolates resistant to moxifloxacin were also resistant to clindamycin. Ribotype 017 had a quite different pattern of resistance. For clindamycin and moxifloxacin, the resistance rate of ribotype 017 was 100% and 95.0%, respectively, and 95.0% were also resistant to rifaximin.
Sixty-four isolates (48.9%) were positive for ermB. The frequency of ermB positivity was related to resistance to clindamycin (63/89; 70.8%) and high-level resistance to clindamycin (52/74; 70.3%) (P < 0.0001 and P < 0.0001, respectively). Frequencies of ermB positivity varied among the PCR ribotypes. Clindamycin resistance rates of ribotypes 018 and 017 were 91.4% and 100%, respectively. Among the clindamycin-resistant ribotype 018 and 017 strains, the ermB gene was detected in 100% and 85.0%, respectively. Ribotypes 001 and 112 had high rates of resistance to clindamycin (84.2% and 100%), but few of the clindamycin-resistant isolates were ermBpositive (5.3% and 25.0%, respectively). 4. Discussion Rates of resistance to clindamycin in C. difficile isolates of 22–88% have been reported, depending on the study setting [11,12]. In the present study, the rate of resistance to clindamycin was high (67.9%) and was similar to Kim et al.’s figure of 60% for South Korea [7]. Previous use of clindamycin had a 4.2-fold higher prevalence odds ratio for acquisition of clindamycin resistance than
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
27
Table 2 Results of susceptibility testing to clindamycin, moxifloxacin and rifaximin according to PCR ribotype. PCR ribotype
Total no.
Clindamycin
Moxifloxacin
MIC (mg/L)
018 017 001 014 002 112 027
35 20 19 7 6 4 3
%R
Range
MIC50
MIC90
2 to >256 16 to >256 1.5 to >256 2–12 2 to >256 12 to >256 3–6
>256 >256 >256 4 3 >256 6
>256 >256 >256 12 >256 >256 6
91.4 100.0 84.2 28.6 16.7 100.0 0.0
Rifaximin
MIC (mg/L)
%R
Range
MIC50
MIC90
2–32 2–64 <0.25–128 0.5–2 2–16 2–128 1–2
16 32 32 2 2 64 2
16 64 128 2 16 128 2
91.4 95.0 78.9 0.0 33.3 75.0 0.0
MIC (mg/L)
%R
Range
MIC50
MIC90
<0.003 to >8 0.007 to >8 <0.003–2 0.007–0.015 0.003–0.015 <0.003–0.007 0.015
0.007 >8 0.007 0.007 0.007 0.007 0.015
0.015 >8 2 0.015 0.015 0.007 0.015
5.7 95.0 0.0 0.0 0.0 0.0 0.0
PCR, polymerase chain reaction; MIC, minimum inhibitory concentration; MIC50/90 , MIC for 50% and 90% of the isolates respectively; %R, percent resistant.
for the other antimicrobial agents [13]. However, Solomon et al. reported that exposure to macrolide–lincosamide–streptogramin B was not associated with acquisition of resistance or presence of the ermB gene [11]. Previous exposure to clindamycin was marginally associated with resistance to clindamycin (P = 0.053). The rate of resistance to moxifloxacin in this study was 62.6%, higher than the 42% reported by Kim et al. for South Korea [7], but similar to that reported by Huang et al. for China (61.8%) [12]. Resistance to moxifloxacin has received attention because of the outbreak involving BI/NAP1 C. difficile isolates [4]. Historically BI/NAP1 strains were not resistant to moxifloxacin; however, with the increasing use of fluoroquinolones, all isolates of the epidemic strain are now resistant to moxifloxacin. Since the introduction of new fluoroquinolones, the rate of resistance to moxifloxacin has increased [14,15]; it was 10% in 1985–2001 and 56% in 2002–2008, and the overall rate was 38.4–87.1% [11,16]. Estimates of the frequency of resistance of C. difficile to rifamycin have varied between 3% and 53.5% in different geographic locations. As there is no recommended breakpoint for rifaximin, many workers have used the breakpoint for rifampicin, and the rate of resistance to rifaximin is seen as like that to rifampicin [12]. In the case of rifaximin, the frequency of resistance was reported as 2.7–4.7% in the USA and 29.1% in China [17,18]. In the current study, the 19.1% resistance rate to rifaximin was similar to that in the Chinese study [12]. Previous use of rifamycin had a relative risk of 2.4 for developing infection due to rifampicin-resistant C. difficile [19]. Only 2 (1.5%) of the 131 HA-CDI patients in the current study had histories of prior exposure to rifamycin, and the isolates from these patients were resistant to rifaximin, with MICs of >8 mg/L. The number of patients who had been exposed to rifamycin was too small for statistical analysis, therefore more cases are needed to elucidate the risk factors for rifaximin resistance. To compare resistance rates to rifaximin in Asia and the USA, nationwide patterns of rifamycin prescription should be surveyed. One unusual strain resistant to metronidazole and vancomycin was encountered. Recent studies have reported decreased rates of susceptibility to metronidazole and vancomycin [20]. According to European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines (v.2.0; http://www.eucast.org), MIC breakpoints of 2 mg/L define resistance to metronidazole and vancomycin. By these criteria, seven metronidazole-resistant strains (5.3%) were isolated, comprising three ribotype 017 strains, two ribotype 018, one ribotype 001 and one strain of undetermined ribotype, as well as four vancomycin-resistant strains (3.1%), comprising two ribotype 018 strains, 1 ribotype 017 and 1 ribotype 002 (Fig. 1). None of the patients harbouring the metronidazoleresistant strains had been exposed to metronidazole within the previous 2 months. Among them, three cases were treated with metronidazole without recurrence. One case did not improve CDI with metronidazole on treatment day 3 and after change to vancomycin resolved diarrhoea without recurrence.
The remaining three cases did not receive medication because of self-resolved diarrhoea. Of the four patients yielding vancomycinresistant strains, two had been exposed to vancomycin within the previous 2 months. The treatment choice of the four cases was metronidazole. There was neither treatment failure nor recurrence among them. There are no available data for comparing the resistance rates for metronidazole and vancomycin according to the EUCAST criteria. Rates of resistance to these antimicrobial agents could be increased by adjusting the EUCAST criteria. The pattern of resistance of C. difficile to antimicrobial agents differed depending on PCR ribotype. Ribotype 018 was the most common type in this study and is reported to be emerging in Europe, especially in Italy [21]. There is little published information about the antibiotic susceptibility of this ribotype. In the study by Spigaglia et al., its rate of resistance to fluoroquinolones was 56%, and these workers suggested that the increased rate of resistance to fluoroquinolones of this ribotype was responsible for its increased prevalence [15]. The rate of resistance of 018 strains to moxifloxacin (91.4%) was higher than the average rate of the 131 isolates in the current study (62.6%). On the basis of the EUCAST breakpoints, the rates of resistance of ribotype 018 to metronidazole and vancomycin were both 5.7% (2/35). Ribotype 001 was the second most common type in Europe in 2008 [21] and the third most common type in the current study. The rates of resistance of ribotype 001 to clindamycin varied from 0% to 85.4% [11,22] and the rates of resistance to moxifloxacin were reported to vary from 50% to 98.9%. Of the ribotype 001 isolates from Scotland, 24.4% showed reduced susceptibility to metronidazole (MICs ≥ 6 mg/L) [20]. In the current study, rates of resistance of ribotype 001 to clindamycin and moxifloxacin were as high as 84.2% and 78.9%, respectively. There has been no previous report about the rifaximin resistance of ribotype 001, and none of the strains in this study were resistant to rifaximin. The metronidazole resistance rate based on EUCAST criteria was 5.3% (1/19). Ribotype 014 was the fourth most common type in this study. Its reported resistance rates to clindamycin were 0–27.6% [11,22,23]. One study reported a resistance rate to clindamycin as 100% [14]. The rate of resistance to moxifloxacin was <10% [14,23]. In the current sample, ribotype 014 also had a low rate of resistance to clindamycin (28.6%) and there were no resistance to moxifloxacin or rifaximin. Ribotype 002 was the fifth most common type in this study. Its resistance rate to clindamycin has been reported variously as 2.4–92% [14,22] and a previous study reported no isolate resistant to moxifloxacin [23]. The MICs of vancomycin for all three ribotype 002 isolates exceeded 4 mg/L [24]. In the current study, the resistance rates of ribotype 002 to clindamycin and moxifloxacin were low (16.7% and 33.3%) and only one isolate had a vancomycin MIC of 4 mg/L (data not shown). The historic BI/NAP1 strain (ribotype 027) was not resistant to moxifloxacin; however, with the increasing use of
28
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
fluoroquinolones, all healthcare-associated isolates of the epidemic BI/NAP1 strain are now resistant to moxifloxacin [4]. Their rate of resistance to clindamycin was relatively low (0–20%) even though resistance to ciprofloxacin approached 100% [11,23,25]. Moxifloxacin was introduced in Hanyang University Hospital in 2004. The resistance rate to moxifloxacin (62.6%) and the exposure rate to the same class of antibiotics (45.0%) showed a significant correlation in this study (P = 0.019). However, there has been no reported outbreak caused by the BI/NAP1/027 strain in South Korea [26] and there were no moxifloxacin- or clindamycinresistant strains among the ribotype 027 isolates in the current study. Some ribotype 017 strains were found to be multidrug-resistant in other studies [25]. Some had high-level resistance to clindamycin and harboured the ermB gene [27]. The rate of resistance to clindamycin of ribotype 017 was reported to be 90–100% and the same was true for clindamycin in the current study [23,25]. There has been one study on the rate of resistance of ribotype 017 to rifamycins, and the reported rates for rifaximin and rifampicin were 75.6% and 96.9%, respectively [12]. In the present study, the resistance rate to rifaximin of ribotype 017 was also 95.0%, and all the resistant isolates in this study harboured the rpoB mutations H502N and R505K as the basis of their resistance. Using the EUCAST breakpoint, the metronidazole resistance rate of ribotype 017 was 15% (3/20) and that of vancomycin was 5% (1/20) (data not shown). One of the isolates resistant to vancomycin was also resistant to metronidazole. Good concordance between possession of the ermB gene and combined clindamycin resistance has been reported [28]. There was a high frequency of the ermB gene among clindamycinresistant strains of ribotypes 001 and 017 (85.8–88% and 89.9–94.4%, respectively) [29,30]. In the current study, possession of the ermB gene was highly correlated with resistance to clindamycin (P < 0.0001). Among the clindamycin-resistant strains, the ermB gene was 100% positive in ribotype 018 and 85.0% positive in ribotype 017.There was only one ermB gene-positive strain among the ribotype 001 isolates. A significant association between clindamycin resistance and possession of the ermB gene was found only in ribotype 018 isolates (P < 0.0001) (data not shown). Other resistance mechanisms should be considered to account for the high rate of resistance to clindamycin among C. difficile isolates, especially for ribotypes 001 and 112. Although there was no association between previous exposure to clindamycin and possession of the ermB gene (P = 0.944) (data not shown), previous use of clindamycin was correlated with resistance to clindamycin among the ermB gene-negative C. difficile isolates (P = 0.012) (data not shown). This finding suggests that previous exposure to clindamycin does not influence the presence/absence of the ermB gene but could be an important factor underlying resistance in ermB gene-negative C. difficile isolates. In conclusion, resistance rates to clindamycin, moxifloxacin and rifaximin differ depending on the PCR ribotype. Ribotype 018 has a high rate of resistance to clindamycin and moxifloxacin but a low rate of resistance to rifaximin. Many ribotype 017 isolates were multiply resistant to clindamycin, moxifloxacin and rifaximin. None of the strains encountered were resistant to metronidazole, vancomycin or TZP. Previous use of moxifloxacin influenced the likelihood of acquisition of resistance to that drug, and possession of the ermB gene was related to resistance to clindamycin.
Acknowledgments The moxifloxacin Etest strips were kindly supplied by bioMérieux SA (Lyon, France) and rifaximin was donated by Alfa Wassermann (Pescara, Italy).
Funding: This work was supported by the research fund of Hanyang University (Seoul, South Korea) (HY-2010-MC). Competing interests: None declared. Ethical approval: This study was approved by the institutional review board of Hanyang University Hospital (Seoul, South Korea) (HYUH IRB 2010-R-12). Informed consent was waived by the board.
References [1] Pepin J, Valiquette L, Alary ME, Villemure P, Pelletier A, Forget K, et al. Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004;171:466–72. [2] Kelly CP, LaMont JT. Clostridium difficile—more difficult than ever. N Engl J Med 2008;359:1932–40. [3] Shin BM, Kuak EY, Yoo HM, Kim EC, Lee K, Kang JO, et al. Multicentre study of the prevalence of toxigenic Clostridium difficile in Korea: results of a retrospective study 2000–2005. J Med Microbiol 2008;57:697–701. [4] McDonald LC, Killgore GE, Thompson A, Owens Jr RC, Kazakova SV, Sambol SP, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005;353:2433–41. [5] Razavi B, Apisarnthanarak A, Mundy LM. Clostridium difficile: emergence of hypervirulence and fluoroquinolone resistance. Infection 2007;35:300–7. [6] Bidet P, Barbut F, Lalande V, Burghoffer B, Petit JC. Development of a new PCR-ribotyping method for Clostridium difficile based on ribosomal RNA gene sequencing. FEMS Microbiol Lett 1999;175:261–6. [7] Kim H, Jeong SH, Roh KH, Hong SG, Kim JW, Shin MG, et al. Investigation of toxin gene diversity, molecular epidemiology, and antimicrobial resistance of Clostridium difficile isolated from 12 hospitals in South Korea. Korean J Lab Med 2010;30:491–7. [8] Clinical and Laboratory Standards Institute. Methods for antimicrobial susceptibility testing of anaerobic bacteria; approved standard. 7th ed. Document M11-A7. Wayne, PA: CLSI; 2007. [9] Huang H, Wu S, Wang M, Zhang Y, Fang H, Palmgren AC, et al. Clostridium difficile infections in a Shanghai hospital: antimicrobial resistance, toxin profiles and ribotypes. Int J Antimicrob Agents 2009;33:339–42. [10] Johnson S, Samore MH, Farrow KA, Killgore GE, Tenover FC, Lyras D, et al. Epidemics of diarrhea caused by a clindamycin-resistant strain of Clostridium difficile in four hospitals. N Engl J Med 1999;341:1645–51. [11] Solomon K, Fanning S, McDermott S, Murray S, Scott L, Martin A, et al. PCR ribotype prevalence and molecular basis of macrolide–lincosamide–streptogramin B (MLSB ) and fluoroquinolone resistance in Irish clinical Clostridium difficile isolates. J Antimicrob Chemother 2011;66:1976–82. [12] Huang H, Weintraub A, Fang H, Wu S, Zhang Y, Nord CE. Antimicrobial susceptibility and heteroresistance in Chinese Clostridium difficile strains. Anaerobe 2010;16:633–5. [13] Samore MH, Venkataraman L, DeGirolami PC, Merrigan MM, Johnson S, Gerding DN, et al. Genotypic and phenotypic analysis of Clostridium difficile correlated with previous antibiotic exposure. Microb Drug Resist 2006;12:23–8. [14] Taori SK, Hall V, Poxton IR. Changes in antibiotic susceptibility and ribotypes in Clostridium difficile isolates from southern Scotland, 1979–2004. J Med Microbiol 2010;59:338–44. [15] Spigaglia P, Barbanti F, Dionisi AM, Mastrantonio P. Clostridium difficile isolates resistant to fluoroquinolones in Italy: emergence of PCR ribotype 018. J Clin Microbiol 2010;48:2892–6. [16] Pituch H, Obuch-Woszczatynski P, Wultanska D, Nurzynska G, Harmanus C, Banaszkiewicz A, et al. Characterization and antimicrobial susceptibility of Clostridium difficile strains isolated from adult patients with diarrhoea hospitalized in two university hospitals in Poland, 2004–2006. J Med Microbiol 2011;60:1200–5. [17] Hecht DW, Galang MA, Sambol SP, Osmolski JR, Johnson S, Gerding DN. In vitro activities of 15 antimicrobial agents against 110 toxigenic Clostridium difficile clinical isolates collected from 1983 to 2004. Antimicrob Agents Chemother 2007;51:2716–19. [18] Jiang ZD, DuPont HL, La Rocco M, Garey KW. In vitro susceptibility of Clostridium difficile to rifaximin and rifampin in 359 consecutive isolates at a university hospital in Houston, Texas. J Clin Pathol 2010;63:355–8. [19] Curry SR, Marsh JW, Shutt KA, Muto CA, O’Leary MM, Saul MI, et al. High frequency of rifampin resistance identified in an epidemic Clostridium difficile clone from a large teaching hospital. Clin Infect Dis 2009;48: 425–9. [20] Baines SD, O’Connor R, Freeman J, Fawley WN, Harmanus C, Mastrantonio P, et al. Emergence of reduced susceptibility to metronidazole in Clostridium difficile. J Antimicrob Chemother 2008;62:1046–52. [21] Bauer MP, Notermans DW, van Benthem BH, Brazier JS, Wilcox MH, Rupnik M, et al. Clostridium difficile infection in Europe: a hospital-based survey. Lancet 2011;377:63–73. [22] John R, Brazier JS. Antimicrobial susceptibility of polymerase chain reaction ribotypes of Clostridium difficile commonly isolated from symptomatic hospital patients in the UK. J Hosp Infect 2005;61:11–14. [23] Barbut F, Mastrantonio P, Delmee M, Brazier J, Kuijper E, Poxton I. Prospective study of Clostridium difficile infections in Europe with phenotypic and genotypic characterisation of the isolates. Clin Microbiol Infect 2007;13:1048–57.
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29 [24] Noren T, Alriksson I, Akerlund T, Burman LG, Unemo M. In vitro susceptibility to 17 antimicrobials of clinical Clostridium difficile isolates collected in 1993–2007 in Sweden. Clin Microbiol Infect 2010;16:1104–10. [25] Goorhuis A, Debast SB, Dutilh JC, van Kinschot CM, Harmanus C, Cannegieter SC, et al. Type-specific risk factors and outcome in an outbreak with 2 different Clostridium difficile types simultaneously in 1 hospital. Clin Infect Dis 2011;53:860–9. [26] Kim J, Pai H, Seo MR, Kang JO. Epidemiology and clinical characteristics of Clostridium difficile infection in a Korean tertiary hospital. J Korean Med Sci 2011;26:1258–64. [27] Bendle JS, James PA, Bennett PM, Avison MB, Macgowan AP, Al-Shafi KM. Resistance determinants in strains of Clostridium difficile from two geographically distinct populations. Int J Antimicrob Agents 2004;24:619–21.
29
[28] Pituch H, van Leeuwen W, Maquelin K, Wultanska D, Obuch-Woszczatynski P, Nurzynska G, et al. Toxin profiles and resistances to macrolides and newer fluoroquinolones as epidemicity determinants of clinical isolates of Clostridium difficile from Warsaw, Poland. J Clin Microbiol 2007;45:1607–10. [29] Goorhuis A, Legaria MC, van den Berg RJ, Harmanus C, Klaassen CH, Brazier JS, et al. Application of multiple-locus variable-number tandem-repeat analysis to determine clonal spread of toxin A-negative Clostridium difficile in a general hospital in Buenos Aires, Argentina. Clin Microbiol Infect 2009;15:1080–6. [30] Pituch H, Brazier JS, Obuch-Woszczatynski P, Wultanska D, MeiselMikolajczyk F, Luczak M. Prevalence and association of PCR ribotypes of Clostridium difficile isolated from symptomatic patients from Warsaw with macrolide–lincosamide–streptogramin B (MLSB ) type resistance. J Med Microbiol 2006;55:207–13.
Contents lists available at SciVerse ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Association between PCR ribotypes and antimicrobial susceptibility among Clostridium difficile isolates from healthcare-associated infections in South Korea Jieun Kim a , Jung Oak Kang b,∗ , Hyunjoo Pai a , Tae Yeal Choi b a b
Department of Internal Medicine, Hanyang University College of Medicine, Seoul, South Korea Department of Laboratory Medicine, Hanyang University College of Medicine, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 2 March 2012 Accepted 22 March 2012 Keywords: Clostridium difficile Antimicrobial susceptibility Ribotyping
a b s t r a c t In this study, the association between antimicrobial susceptibility, PCR ribotype and presence of the ermB gene in clinical isolates of Clostridium difficile was investigated. PCR ribotyping and ermB gene PCR were performed on 131 C. difficile isolates. The susceptibility of these isolates to metronidazole, vancomycin, piperacillin/tazobactam (TZP), clindamycin, moxifloxacin and rifaximin was also determined. Use of antibiotics within the previous 2 months was documented. Resistance rates to clindamycin, moxifloxacin and rifaximin were 67.9%, 62.6% and 19.1%, respectively. No metronidazole, vancomycin or TZP resistance was detected. Previous exposure to moxifloxacin was significantly correlated with resistance to this antibiotic, but prior use of clindamycin was not significantly correlated with clindamycin resistance. Sixty-four strains (48.9%) carried the ermB gene, of which all but one (98.5%) were resistant to clindamycin. The clindamycin resistance rates of the common PCR ribotypes (018, 017 and 001) were 91.4%, 100% and 84.2%, respectively, and their moxifloxacin resistance rates were 91.4%, 95.0% and 78.9%, respectively. Resistance rates to rifaximin were 5.7% and 95.0% in ribotype 018 and 017 strains, whilst none of the 001 strains were resistant to rifaximin. In conclusion, the common ribotypes 018, 017 and 001 of C. difficile have high rates of resistance to clindamycin and moxifloxacin, but differ greatly in the frequency of rifaximin resistance. © 2012 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction
both [4]. The most commonly used antimicrobial in the 1970s was clindamycin, and in the 1980s it was cephalosporins, but recent epidemic cases in Canada suggest that fluoroquinolones may now play an important role [5]. Therefore, surveillance of prior antimicrobial use, antimicrobial resistance patterns in the community and hospitals, and characterisation of the corresponding molecular mechanisms are very important for avoiding treatment failure or complications as well as preventing nosocomial or community spread of epidemic strains. In this study, the association between recent antimicrobial use and antimicrobial resistance of clinical isolates of toxigenic C. difficile in a tertiary care hospital in South Korea was investigated. Antimicrobial resistance in relation to epidemic PCR ribotype patterns and resistance genes was also analysed.
Clostridium difficile has been known for more than 30 years as the primary cause of antibiotic-associated diarrhoea and pseudomembranous colitis. Currently, a new strain known as type BI from restriction endonuclease analysis, North American pulsed-field type 1 (NAP1) from pulsed-field gel electrophoresis and ribotype 027 from PCR ribotyping has become the single most important epidemic strain causing C. difficile infection (CDI) worldwide [1]. Because of the ribotype 027 strain epidemic, the incidence of CDI, the proportion of complicated cases and fatality rates have all increased in North America and Europe [2]. However, the ribotype 027 strain was rarely reported in South Korea until recently, and molecular epidemiology has instead revealed a nationwide epidemic of A− B+ strains of C. difficile starting in 2003 [3]. There are several explanations for the changing epidemiology of CDI, including alterations of antimicrobial use and the emergence of a new strain with increased virulence, antimicrobial resistance or
2.1. Setting and study design
∗ Corresponding author. Present address: Department of Laboratory Medicine, Hanyang University, Guri Hospital, Gyeongchun-ro 153, Guri City, Gyeonggi-do, 471-701, South Korea. Tel.: +82 31 560 2572; fax: +82 31 560 2585. E-mail address: [email protected] (J.O. Kang).
This study was conducted at Hanyang University Hospital, a 900-bed tertiary care facility located in Seoul, South Korea. From September 2008 through January 2010, all C. difficile isolates from patients older than 18 years of age with healthcare-associated C. difficile infections (HA-CDIs) were enrolled. The study was approved
2. Materials and methods
0924-8579/$ – see front matter © 2012 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. http://dx.doi.org/10.1016/j.ijantimicag.2012.03.015
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
by the institutional review board of Hanyang University Hospital (HYUH IRB 2010-R-12). Informed consent was waived by the board. 2.2. Definitions HA-CDI was defined when the C. difficile isolates from stool cultures were positive for toxin genes (tcdA, tcdB, cdtA or cdtB) by multiplex PCR or gave positive results in toxin antigen assay A&B using a commercial kit (VIDAS® C. difficile Toxins A & B; bioMérieux SA, Marcy l’Étoile, France) in patients with diarrhoea that developed ≥72 h after hospitalisation or within 2 months of the last discharge, provided the patient had not resided in a healthcare facility. Previous use of antimicrobial agents was defined when a patient had used the same class of antibiotics for ≥3 days within the last 2 months. Use of clindamycin, fluoroquinolones (moxifloxacin, levofloxacin and ciprofloxacin), rifamycins (rifampicin and rifaximin), metronidazole, glycopeptides (vancomycin and teicoplanin) and -lactam/-lactamase inhibitors [piperacillin/tazobactam (TZP), ampicillin/sulbactam and amoxicillin/clavulanic acid] was investigated. 2.3. Isolation of Clostridium difficile
25
Resistance breakpoints were as defined by the Clinical and Laboratory Standards Institute (CLSI) [8] (Table 1). For those antimicrobial agents with no defined breakpoint, resistance was defined according to a previous report by Huang et al. [9] (Table 1). Highlevel resistance was defined as MICs for clindamycin of ≥256 mg/L and for moxifloxacin of ≥32 mg/L [9]. 2.6. Detection of the ermB gene ermB gene-specific PCR was performed with template DNA, as described previously, with primers 2980 (5 AATTAAGTAAACAGGTAACGTT3 ) and 2981 (5 GCTCCTTGGAAGCTGTCAGTAG3 ) [10]. A PCR product of 688 bp on electrophoresis was considered a positive result for ermB. 2.7. Statistical methods SPSS v.13.0 for Windows (SPSS Inc., Chicago, IL) was used for statistical analysis. Categorical variables were analysed by Pearson’s 2 test or Fisher’s exact test. A P-value of <0.05 by two-tailed test was considered statistically significant. 3. Results
Stool specimens were cultured anaerobically on cycloserine– cefoxitin–taurocholate agar (Oxoid Ltd., Cambridge, UK) supplemented with 7% horse blood after alcohol shock treatment of stool specimens. Colonies of C. difficile were identified with an API® Rapid ID 32A system (bioMérieux SA, Lyon, France). 2.4. PCR ribotyping of Clostridium difficile PCR ribotyping was performed using template DNA as described elsewhere [6]. Amplification products were fractionated by electrophoresis through 8% acrylamide gels (30% acrylamide/bis solution, 29:1; Bio-Rad, Richmond, CA) and were visualised under ultraviolet by staining gels for 30 min with ethidium bromide. Banding patterns were checked visually. Each unique pattern was assigned its own ribotype code and was matched with the PCR ribotypes of the reference strains ribotype 027 and ATCC 43598 strain (ribotype 017) [7]. 2.5. Antimicrobial susceptibility tests Susceptibility tests were conducted on 143 isolates from 188 HA-CDI patients, and minimum inhibitory concentrations (MICs) of six antimicrobial agents (metronidazole, vancomycin, TZP, clindamycin, moxifloxacin and rifaximin) were determined for 131 C. difficile isolates from 131 HA-CDI patients. The MICs of clindamycin, moxifloxacin, metronidazole, vancomycin and TZP were determined by Etest (AB-BIODISK, Solna, Sweden) and those of rifaximin were determined by the agar dilution test. Clostridium difficile ATCC 700057 was used as a quality control strain for the susceptibility tests.
Among 131 isolates, 20 toxin B-only, 100 toxin A and B, and 11 binary toxin as well as toxin A and B-producing isolates were included. 3.1. Minimum inhibitory concentration distribution of Clostridium difficile for various antimicrobial agents The MIC distribution patterns of the 131 C. difficile isolates differed depending on the antimicrobial agent (Fig. 1). The MICs of clindamycin (Fig. 1A), moxifloxacin (Fig. 1B) and rifaximin (Fig. 1C) each formed bimodal distributions clearly separated by their breakpoints. The distributions of the MICs for TZP (Fig. 1D), metronidazole (Fig. 1E) and vancomycin (Fig. 1F) were unimodal with narrow ranges of MICs. None of the isolates were resistant to metronidazole, vancomycin or TZP (Table 1). The rates of resistance to clindamycin and moxifloxacin were 67.9% and 62.6%, respectively, and 80 (61.1%) of the isolates were resistant to both drugs. The rates of high-level resistance to clindamycin (≥256 mg/L) and moxifloxacin (≥32 mg/L) were 56.5% and 29.0%, respectively. Twenty-five isolates (19.1%) were resistant to rifaximin, of which 22 (88.0%) were also resistant to clindamycin and 23 (92.0%) to moxifloxacin, with close correlations between the resistances (P = 0.017 and P < 0.0001, respectively). 3.2. Correlation between antibiotic resistance and previous antibiotic exposure Previous use of clindamycin had a marginal effect on the proportion of strains resistant to clindamycin (P = 0.053). On the other
Table 1 Antimicrobial resistance rates of 131 isolates from healthcare-associated Clostridium difficile infection and previous antibiotic history. Antibiotic
Breakpoint (mg/L)a
Resistance rate [n (%)]
Exposure to same class of antibiotics [n (%)]
P-value
Clindamycin Moxifloxacin Rifaximin Metronidazole Vancomycin Piperacillin/tazobactam
≥8 ≥8 ≥4 ≥32 ≥32 ≥128/4
89 (67.9) 82 (62.6) 25 (19.1) 0 (0.0) 0 (0.0) 0 (0.0)
29 (22.1) 59 (45.0) 2 (1.5) 35 (26.7) 27 (20.6) 41 (31.3)
0.053 0.019
a The breakpoints for most antimicrobial agents were considered according to Clinical and Laboratory Standards Institute guidelines [8]. Resistance to rifaximin and vancomycin was considered according to a previous report by Huang et al. [9].
26
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
Fig. 1. Minimum inhibitory concentration (MIC) distribution of 131 isolates of Clostridium difficile to various antimicrobial agents. (A–C) MICs of clindamycin (A), moxifloxacin (B) and rifaximin (C), separated by vertical grey line indicating the breakpoint of each agent. (D–F) Piperacillin/tazobactam (D), metronidazole (E) and vancomycin (F) showed MICs with a unimodal distribution.
hand, previous use of fluoroquinolones was highly related to the frequency of resistance to moxifloxacin (P = 0.019). In the case of rifaximin, none of the patients had previous exposure to rifaximin and only two had been exposed to rifampicin, and these latter were resistant to rifaximin. Previous use of metronidazole, vancomycin and TZP was not associated with resistance.
Eleven binary toxin-producing strains were identified, including three of ribotype 027. Ribotype 027 was not resistant to any of the six antimicrobial agents. None of the other binary toxin-producing isolates was resistant to any of the six antimicrobial agents, except for one with a MIC of 8 mg/L for rifaximin. 3.4. Presence of the ermB gene
3.3. Antibiotic resistance according to PCR ribotypes of Clostridium difficile A total of 35 distinct PCR ribotypes was observed among the 131 isolates, the most common being ribotype 018 (35; 26.7%), followed by ribotype 017 (20; 15.3%) and ribotype 001 (19; 14.5%). MICs for clindamycin, moxifloxacin and rifaximin among the isolates of seven ribotypes are shown in Table 2. The most common ribotype (018) had high rates of resistance to clindamycin (91.4%) and moxifloxacin (91.4%), but a low rate of resistance to rifaximin (5.7%). All of the clindamycin-resistant 018 isolates were also resistant to moxifloxacin. Ribotypes 001 and 112 also had high rates of resistance to clindamycin (84.2% and 100%) and moxifloxacin (78.9% and 75.0%), and none of the isolates was resistant to rifaximin. All of the 001 and 112 isolates resistant to moxifloxacin were also resistant to clindamycin. Ribotype 017 had a quite different pattern of resistance. For clindamycin and moxifloxacin, the resistance rate of ribotype 017 was 100% and 95.0%, respectively, and 95.0% were also resistant to rifaximin.
Sixty-four isolates (48.9%) were positive for ermB. The frequency of ermB positivity was related to resistance to clindamycin (63/89; 70.8%) and high-level resistance to clindamycin (52/74; 70.3%) (P < 0.0001 and P < 0.0001, respectively). Frequencies of ermB positivity varied among the PCR ribotypes. Clindamycin resistance rates of ribotypes 018 and 017 were 91.4% and 100%, respectively. Among the clindamycin-resistant ribotype 018 and 017 strains, the ermB gene was detected in 100% and 85.0%, respectively. Ribotypes 001 and 112 had high rates of resistance to clindamycin (84.2% and 100%), but few of the clindamycin-resistant isolates were ermBpositive (5.3% and 25.0%, respectively). 4. Discussion Rates of resistance to clindamycin in C. difficile isolates of 22–88% have been reported, depending on the study setting [11,12]. In the present study, the rate of resistance to clindamycin was high (67.9%) and was similar to Kim et al.’s figure of 60% for South Korea [7]. Previous use of clindamycin had a 4.2-fold higher prevalence odds ratio for acquisition of clindamycin resistance than
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
27
Table 2 Results of susceptibility testing to clindamycin, moxifloxacin and rifaximin according to PCR ribotype. PCR ribotype
Total no.
Clindamycin
Moxifloxacin
MIC (mg/L)
018 017 001 014 002 112 027
35 20 19 7 6 4 3
%R
Range
MIC50
MIC90
2 to >256 16 to >256 1.5 to >256 2–12 2 to >256 12 to >256 3–6
>256 >256 >256 4 3 >256 6
>256 >256 >256 12 >256 >256 6
91.4 100.0 84.2 28.6 16.7 100.0 0.0
Rifaximin
MIC (mg/L)
%R
Range
MIC50
MIC90
2–32 2–64 <0.25–128 0.5–2 2–16 2–128 1–2
16 32 32 2 2 64 2
16 64 128 2 16 128 2
91.4 95.0 78.9 0.0 33.3 75.0 0.0
MIC (mg/L)
%R
Range
MIC50
MIC90
<0.003 to >8 0.007 to >8 <0.003–2 0.007–0.015 0.003–0.015 <0.003–0.007 0.015
0.007 >8 0.007 0.007 0.007 0.007 0.015
0.015 >8 2 0.015 0.015 0.007 0.015
5.7 95.0 0.0 0.0 0.0 0.0 0.0
PCR, polymerase chain reaction; MIC, minimum inhibitory concentration; MIC50/90 , MIC for 50% and 90% of the isolates respectively; %R, percent resistant.
for the other antimicrobial agents [13]. However, Solomon et al. reported that exposure to macrolide–lincosamide–streptogramin B was not associated with acquisition of resistance or presence of the ermB gene [11]. Previous exposure to clindamycin was marginally associated with resistance to clindamycin (P = 0.053). The rate of resistance to moxifloxacin in this study was 62.6%, higher than the 42% reported by Kim et al. for South Korea [7], but similar to that reported by Huang et al. for China (61.8%) [12]. Resistance to moxifloxacin has received attention because of the outbreak involving BI/NAP1 C. difficile isolates [4]. Historically BI/NAP1 strains were not resistant to moxifloxacin; however, with the increasing use of fluoroquinolones, all isolates of the epidemic strain are now resistant to moxifloxacin. Since the introduction of new fluoroquinolones, the rate of resistance to moxifloxacin has increased [14,15]; it was 10% in 1985–2001 and 56% in 2002–2008, and the overall rate was 38.4–87.1% [11,16]. Estimates of the frequency of resistance of C. difficile to rifamycin have varied between 3% and 53.5% in different geographic locations. As there is no recommended breakpoint for rifaximin, many workers have used the breakpoint for rifampicin, and the rate of resistance to rifaximin is seen as like that to rifampicin [12]. In the case of rifaximin, the frequency of resistance was reported as 2.7–4.7% in the USA and 29.1% in China [17,18]. In the current study, the 19.1% resistance rate to rifaximin was similar to that in the Chinese study [12]. Previous use of rifamycin had a relative risk of 2.4 for developing infection due to rifampicin-resistant C. difficile [19]. Only 2 (1.5%) of the 131 HA-CDI patients in the current study had histories of prior exposure to rifamycin, and the isolates from these patients were resistant to rifaximin, with MICs of >8 mg/L. The number of patients who had been exposed to rifamycin was too small for statistical analysis, therefore more cases are needed to elucidate the risk factors for rifaximin resistance. To compare resistance rates to rifaximin in Asia and the USA, nationwide patterns of rifamycin prescription should be surveyed. One unusual strain resistant to metronidazole and vancomycin was encountered. Recent studies have reported decreased rates of susceptibility to metronidazole and vancomycin [20]. According to European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines (v.2.0; http://www.eucast.org), MIC breakpoints of 2 mg/L define resistance to metronidazole and vancomycin. By these criteria, seven metronidazole-resistant strains (5.3%) were isolated, comprising three ribotype 017 strains, two ribotype 018, one ribotype 001 and one strain of undetermined ribotype, as well as four vancomycin-resistant strains (3.1%), comprising two ribotype 018 strains, 1 ribotype 017 and 1 ribotype 002 (Fig. 1). None of the patients harbouring the metronidazoleresistant strains had been exposed to metronidazole within the previous 2 months. Among them, three cases were treated with metronidazole without recurrence. One case did not improve CDI with metronidazole on treatment day 3 and after change to vancomycin resolved diarrhoea without recurrence.
The remaining three cases did not receive medication because of self-resolved diarrhoea. Of the four patients yielding vancomycinresistant strains, two had been exposed to vancomycin within the previous 2 months. The treatment choice of the four cases was metronidazole. There was neither treatment failure nor recurrence among them. There are no available data for comparing the resistance rates for metronidazole and vancomycin according to the EUCAST criteria. Rates of resistance to these antimicrobial agents could be increased by adjusting the EUCAST criteria. The pattern of resistance of C. difficile to antimicrobial agents differed depending on PCR ribotype. Ribotype 018 was the most common type in this study and is reported to be emerging in Europe, especially in Italy [21]. There is little published information about the antibiotic susceptibility of this ribotype. In the study by Spigaglia et al., its rate of resistance to fluoroquinolones was 56%, and these workers suggested that the increased rate of resistance to fluoroquinolones of this ribotype was responsible for its increased prevalence [15]. The rate of resistance of 018 strains to moxifloxacin (91.4%) was higher than the average rate of the 131 isolates in the current study (62.6%). On the basis of the EUCAST breakpoints, the rates of resistance of ribotype 018 to metronidazole and vancomycin were both 5.7% (2/35). Ribotype 001 was the second most common type in Europe in 2008 [21] and the third most common type in the current study. The rates of resistance of ribotype 001 to clindamycin varied from 0% to 85.4% [11,22] and the rates of resistance to moxifloxacin were reported to vary from 50% to 98.9%. Of the ribotype 001 isolates from Scotland, 24.4% showed reduced susceptibility to metronidazole (MICs ≥ 6 mg/L) [20]. In the current study, rates of resistance of ribotype 001 to clindamycin and moxifloxacin were as high as 84.2% and 78.9%, respectively. There has been no previous report about the rifaximin resistance of ribotype 001, and none of the strains in this study were resistant to rifaximin. The metronidazole resistance rate based on EUCAST criteria was 5.3% (1/19). Ribotype 014 was the fourth most common type in this study. Its reported resistance rates to clindamycin were 0–27.6% [11,22,23]. One study reported a resistance rate to clindamycin as 100% [14]. The rate of resistance to moxifloxacin was <10% [14,23]. In the current sample, ribotype 014 also had a low rate of resistance to clindamycin (28.6%) and there were no resistance to moxifloxacin or rifaximin. Ribotype 002 was the fifth most common type in this study. Its resistance rate to clindamycin has been reported variously as 2.4–92% [14,22] and a previous study reported no isolate resistant to moxifloxacin [23]. The MICs of vancomycin for all three ribotype 002 isolates exceeded 4 mg/L [24]. In the current study, the resistance rates of ribotype 002 to clindamycin and moxifloxacin were low (16.7% and 33.3%) and only one isolate had a vancomycin MIC of 4 mg/L (data not shown). The historic BI/NAP1 strain (ribotype 027) was not resistant to moxifloxacin; however, with the increasing use of
28
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29
fluoroquinolones, all healthcare-associated isolates of the epidemic BI/NAP1 strain are now resistant to moxifloxacin [4]. Their rate of resistance to clindamycin was relatively low (0–20%) even though resistance to ciprofloxacin approached 100% [11,23,25]. Moxifloxacin was introduced in Hanyang University Hospital in 2004. The resistance rate to moxifloxacin (62.6%) and the exposure rate to the same class of antibiotics (45.0%) showed a significant correlation in this study (P = 0.019). However, there has been no reported outbreak caused by the BI/NAP1/027 strain in South Korea [26] and there were no moxifloxacin- or clindamycinresistant strains among the ribotype 027 isolates in the current study. Some ribotype 017 strains were found to be multidrug-resistant in other studies [25]. Some had high-level resistance to clindamycin and harboured the ermB gene [27]. The rate of resistance to clindamycin of ribotype 017 was reported to be 90–100% and the same was true for clindamycin in the current study [23,25]. There has been one study on the rate of resistance of ribotype 017 to rifamycins, and the reported rates for rifaximin and rifampicin were 75.6% and 96.9%, respectively [12]. In the present study, the resistance rate to rifaximin of ribotype 017 was also 95.0%, and all the resistant isolates in this study harboured the rpoB mutations H502N and R505K as the basis of their resistance. Using the EUCAST breakpoint, the metronidazole resistance rate of ribotype 017 was 15% (3/20) and that of vancomycin was 5% (1/20) (data not shown). One of the isolates resistant to vancomycin was also resistant to metronidazole. Good concordance between possession of the ermB gene and combined clindamycin resistance has been reported [28]. There was a high frequency of the ermB gene among clindamycinresistant strains of ribotypes 001 and 017 (85.8–88% and 89.9–94.4%, respectively) [29,30]. In the current study, possession of the ermB gene was highly correlated with resistance to clindamycin (P < 0.0001). Among the clindamycin-resistant strains, the ermB gene was 100% positive in ribotype 018 and 85.0% positive in ribotype 017.There was only one ermB gene-positive strain among the ribotype 001 isolates. A significant association between clindamycin resistance and possession of the ermB gene was found only in ribotype 018 isolates (P < 0.0001) (data not shown). Other resistance mechanisms should be considered to account for the high rate of resistance to clindamycin among C. difficile isolates, especially for ribotypes 001 and 112. Although there was no association between previous exposure to clindamycin and possession of the ermB gene (P = 0.944) (data not shown), previous use of clindamycin was correlated with resistance to clindamycin among the ermB gene-negative C. difficile isolates (P = 0.012) (data not shown). This finding suggests that previous exposure to clindamycin does not influence the presence/absence of the ermB gene but could be an important factor underlying resistance in ermB gene-negative C. difficile isolates. In conclusion, resistance rates to clindamycin, moxifloxacin and rifaximin differ depending on the PCR ribotype. Ribotype 018 has a high rate of resistance to clindamycin and moxifloxacin but a low rate of resistance to rifaximin. Many ribotype 017 isolates were multiply resistant to clindamycin, moxifloxacin and rifaximin. None of the strains encountered were resistant to metronidazole, vancomycin or TZP. Previous use of moxifloxacin influenced the likelihood of acquisition of resistance to that drug, and possession of the ermB gene was related to resistance to clindamycin.
Acknowledgments The moxifloxacin Etest strips were kindly supplied by bioMérieux SA (Lyon, France) and rifaximin was donated by Alfa Wassermann (Pescara, Italy).
Funding: This work was supported by the research fund of Hanyang University (Seoul, South Korea) (HY-2010-MC). Competing interests: None declared. Ethical approval: This study was approved by the institutional review board of Hanyang University Hospital (Seoul, South Korea) (HYUH IRB 2010-R-12). Informed consent was waived by the board.
References [1] Pepin J, Valiquette L, Alary ME, Villemure P, Pelletier A, Forget K, et al. Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004;171:466–72. [2] Kelly CP, LaMont JT. Clostridium difficile—more difficult than ever. N Engl J Med 2008;359:1932–40. [3] Shin BM, Kuak EY, Yoo HM, Kim EC, Lee K, Kang JO, et al. Multicentre study of the prevalence of toxigenic Clostridium difficile in Korea: results of a retrospective study 2000–2005. J Med Microbiol 2008;57:697–701. [4] McDonald LC, Killgore GE, Thompson A, Owens Jr RC, Kazakova SV, Sambol SP, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005;353:2433–41. [5] Razavi B, Apisarnthanarak A, Mundy LM. Clostridium difficile: emergence of hypervirulence and fluoroquinolone resistance. Infection 2007;35:300–7. [6] Bidet P, Barbut F, Lalande V, Burghoffer B, Petit JC. Development of a new PCR-ribotyping method for Clostridium difficile based on ribosomal RNA gene sequencing. FEMS Microbiol Lett 1999;175:261–6. [7] Kim H, Jeong SH, Roh KH, Hong SG, Kim JW, Shin MG, et al. Investigation of toxin gene diversity, molecular epidemiology, and antimicrobial resistance of Clostridium difficile isolated from 12 hospitals in South Korea. Korean J Lab Med 2010;30:491–7. [8] Clinical and Laboratory Standards Institute. Methods for antimicrobial susceptibility testing of anaerobic bacteria; approved standard. 7th ed. Document M11-A7. Wayne, PA: CLSI; 2007. [9] Huang H, Wu S, Wang M, Zhang Y, Fang H, Palmgren AC, et al. Clostridium difficile infections in a Shanghai hospital: antimicrobial resistance, toxin profiles and ribotypes. Int J Antimicrob Agents 2009;33:339–42. [10] Johnson S, Samore MH, Farrow KA, Killgore GE, Tenover FC, Lyras D, et al. Epidemics of diarrhea caused by a clindamycin-resistant strain of Clostridium difficile in four hospitals. N Engl J Med 1999;341:1645–51. [11] Solomon K, Fanning S, McDermott S, Murray S, Scott L, Martin A, et al. PCR ribotype prevalence and molecular basis of macrolide–lincosamide–streptogramin B (MLSB ) and fluoroquinolone resistance in Irish clinical Clostridium difficile isolates. J Antimicrob Chemother 2011;66:1976–82. [12] Huang H, Weintraub A, Fang H, Wu S, Zhang Y, Nord CE. Antimicrobial susceptibility and heteroresistance in Chinese Clostridium difficile strains. Anaerobe 2010;16:633–5. [13] Samore MH, Venkataraman L, DeGirolami PC, Merrigan MM, Johnson S, Gerding DN, et al. Genotypic and phenotypic analysis of Clostridium difficile correlated with previous antibiotic exposure. Microb Drug Resist 2006;12:23–8. [14] Taori SK, Hall V, Poxton IR. Changes in antibiotic susceptibility and ribotypes in Clostridium difficile isolates from southern Scotland, 1979–2004. J Med Microbiol 2010;59:338–44. [15] Spigaglia P, Barbanti F, Dionisi AM, Mastrantonio P. Clostridium difficile isolates resistant to fluoroquinolones in Italy: emergence of PCR ribotype 018. J Clin Microbiol 2010;48:2892–6. [16] Pituch H, Obuch-Woszczatynski P, Wultanska D, Nurzynska G, Harmanus C, Banaszkiewicz A, et al. Characterization and antimicrobial susceptibility of Clostridium difficile strains isolated from adult patients with diarrhoea hospitalized in two university hospitals in Poland, 2004–2006. J Med Microbiol 2011;60:1200–5. [17] Hecht DW, Galang MA, Sambol SP, Osmolski JR, Johnson S, Gerding DN. In vitro activities of 15 antimicrobial agents against 110 toxigenic Clostridium difficile clinical isolates collected from 1983 to 2004. Antimicrob Agents Chemother 2007;51:2716–19. [18] Jiang ZD, DuPont HL, La Rocco M, Garey KW. In vitro susceptibility of Clostridium difficile to rifaximin and rifampin in 359 consecutive isolates at a university hospital in Houston, Texas. J Clin Pathol 2010;63:355–8. [19] Curry SR, Marsh JW, Shutt KA, Muto CA, O’Leary MM, Saul MI, et al. High frequency of rifampin resistance identified in an epidemic Clostridium difficile clone from a large teaching hospital. Clin Infect Dis 2009;48: 425–9. [20] Baines SD, O’Connor R, Freeman J, Fawley WN, Harmanus C, Mastrantonio P, et al. Emergence of reduced susceptibility to metronidazole in Clostridium difficile. J Antimicrob Chemother 2008;62:1046–52. [21] Bauer MP, Notermans DW, van Benthem BH, Brazier JS, Wilcox MH, Rupnik M, et al. Clostridium difficile infection in Europe: a hospital-based survey. Lancet 2011;377:63–73. [22] John R, Brazier JS. Antimicrobial susceptibility of polymerase chain reaction ribotypes of Clostridium difficile commonly isolated from symptomatic hospital patients in the UK. J Hosp Infect 2005;61:11–14. [23] Barbut F, Mastrantonio P, Delmee M, Brazier J, Kuijper E, Poxton I. Prospective study of Clostridium difficile infections in Europe with phenotypic and genotypic characterisation of the isolates. Clin Microbiol Infect 2007;13:1048–57.
J. Kim et al. / International Journal of Antimicrobial Agents 40 (2012) 24–29 [24] Noren T, Alriksson I, Akerlund T, Burman LG, Unemo M. In vitro susceptibility to 17 antimicrobials of clinical Clostridium difficile isolates collected in 1993–2007 in Sweden. Clin Microbiol Infect 2010;16:1104–10. [25] Goorhuis A, Debast SB, Dutilh JC, van Kinschot CM, Harmanus C, Cannegieter SC, et al. Type-specific risk factors and outcome in an outbreak with 2 different Clostridium difficile types simultaneously in 1 hospital. Clin Infect Dis 2011;53:860–9. [26] Kim J, Pai H, Seo MR, Kang JO. Epidemiology and clinical characteristics of Clostridium difficile infection in a Korean tertiary hospital. J Korean Med Sci 2011;26:1258–64. [27] Bendle JS, James PA, Bennett PM, Avison MB, Macgowan AP, Al-Shafi KM. Resistance determinants in strains of Clostridium difficile from two geographically distinct populations. Int J Antimicrob Agents 2004;24:619–21.
29
[28] Pituch H, van Leeuwen W, Maquelin K, Wultanska D, Obuch-Woszczatynski P, Nurzynska G, et al. Toxin profiles and resistances to macrolides and newer fluoroquinolones as epidemicity determinants of clinical isolates of Clostridium difficile from Warsaw, Poland. J Clin Microbiol 2007;45:1607–10. [29] Goorhuis A, Legaria MC, van den Berg RJ, Harmanus C, Klaassen CH, Brazier JS, et al. Application of multiple-locus variable-number tandem-repeat analysis to determine clonal spread of toxin A-negative Clostridium difficile in a general hospital in Buenos Aires, Argentina. Clin Microbiol Infect 2009;15:1080–6. [30] Pituch H, Brazier JS, Obuch-Woszczatynski P, Wultanska D, MeiselMikolajczyk F, Luczak M. Prevalence and association of PCR ribotypes of Clostridium difficile isolated from symptomatic patients from Warsaw with macrolide–lincosamide–streptogramin B (MLSB ) type resistance. J Med Microbiol 2006;55:207–13.