Distribution of ESBLs, AmpC β-lactamases and carbapenemases among Enterobacteriaceae isolates causing intra-abdominal and urinary tract infections in the Asia-Pacific region during 2008–14: results from the Study for Monitoring Antimicrobial Resistance Trends (SMART)

Objectives

To investigate the antimicrobial resistance and assess the molecular characteristics of β-lactamases (ESBLs, AmpC β-lactamases and carbapenemases) among Enterobacteriaceae isolates that caused intra-abdominal infections (IAIs) in patients hospitalized in the Asia-Pacific region during 2008–14.

Methods

Multiplex PCR was used to detect the specific types of β-lactamase in 2893 isolates with MICs of ertapenem >0.5 mg/L. In-hospital acquisition times for most isolates were also delineated.

Results

Among 2728 (94.3%) isolates proven with β-lactamase production, the rates of non-susceptibility to imipenem were low (average = 7.9%) among IAI Enterobacteriaceae isolates from all Asia-Pacific countries except Vietnam (17.7%) and the Philippines (10.2%). A stepwise and significant increase in annual rates of carbapenemase production among these isolates was noted. CTX-M-15 and CTX-M-14 were the dominant ESBL variants in most IAI Enterobacteriaceae species. The most abundant AmpC β-lactamase variants were bla_CMY-2 among isolates of Escherichia coli and bla_DHA-1 among isolates of Klebsiella pneumoniae. In addition, the IAI Enterobacteriaceae isolates harbouring a bla_CMY-2 or bla_DHA-1 allele were associated with high community-acquired rates (38.0% and 42.6%, respectively). AmpC ACT and MIR variants were mostly detected in Enterobacter species. The bla_NDM-1,4,5,7-harbouring isolates of E. coli, K. pneumoniae and Enterobacter cloacae were most commonly identified among IAI isolates from Vietnam and the Philippines. Also of note, bla_OXA-48-harbouring IAI Enterobacteriaceae isolates were detected exclusively in Vietnam.

Conclusions

The high resistance burden in Vietnam and the Philippines warrants aggressive control policies to combat the worsening trend in antimicrobial resistance among Enterobacteriaceae species causing IAIs.

Introduction

β-Lactamase-producing species of the Enterobacteriaceae family are notoriously resistant to penicillin, extended-spectrum cephalosporins and carbapenems, which severely limits the therapeutic options for treating patients with infections caused by resistant pathogens.^1–3 During the past 10 years, in community settings, there has been a marked increase in isolation rates of MDR Enterobacteriaceae species expressing ESBLs, AmpC β-lactamases or carbapenemases.^2,4 Asia is a known epicentre of antibacterial drug resistance due to ESBL-producing Enterobacteriaceae species, with CTX-M recognized as the most common β-lactamase.^5–7 In the last decade, New Delhi metallo-β-lactamase (NDM)- and Klebsiella pneumoniae carbapenemase (KPC)-producing Enterobacteriaceae have become endemic on the Indian subcontinent and in China.^2,6,7 The prevalence of ESBL-producing and carbapenemase-producing Enterobacteriaceae varies remarkably across the Asia-Pacific region.⁷

The Study for Monitoring Antimicrobial Resistance Trends (SMART) has continuously monitored the in vitro activities of antimicrobial agents against Gram-negative aerobic pathogens that cause intra-abdominal infections (IAIs) since 2002. This study investigated the geographical distribution and produced detailed genetic data regarding production of ESBLs, AmpC β-lactamases and carbapenemases in clinical isolates of Enterobacteriaceae obtained from patients with IAIs in the Asia-Pacific region during 2008–14.

Materials and methods

Study countries and isolates

Thirty-seven medical centres in 12 Asia-Pacific countries/regions participated in the SMART programme during 2008–14, including Australia (n = 5), the Hong Kong Special Administrative Region of China (n = 2), Japan (n = 3), Kazakhstan (n = 1), Malaysia (n = 2), New Zealand (n = 4), Singapore (n = 2), South Korea (n = 2), Taiwan (n = 8), Thailand (n = 2), the Philippines (n = 2) and Vietnam (n = 4). Consecutive isolates of Enterobacteriaceae cultured from clinical specimens of patients with IAIs were collected. Clinical specimens from urine with significant pyuria, tissues of the urinary tract (ureter, urinary bladder, prostate, urethra) and intra-abdominal sites related to IAIs comprised tissue, fluid or deep wound cultures obtained intraoperatively, and fluid from paracentesis or percutaneous aspiration of abscesses. Duplicate isolates (the same species from the same patient within 30 days from the first positive culture) were excluded. Isolates collected within 48 h of admission to hospitals were presumptively categorized as community-acquired (CA) IAIs, while those collected ≥48 h after admission were categorized as hospital-acquired (HA) IAIs.

Antimicrobial susceptibility testing and molecular detection of ESBLs, AmpC β-lactamases and carbapenemases

Antimicrobial susceptibility testing (MICs of cefoxitin, ceftriaxone, piperacillin/tazobactam, cefepime, ertapenem, imipenem, amikacin, ciprofloxacin and levofloxacin) by the broth microdilution method, quality control testing and molecular analyses (via multiplex PCR) for all Enterobacteriaceae isolates were performed by the central laboratory (International Health Management Associates, Inc., Schaumburg, IL, USA). Antimicrobial susceptibility results were interpreted according to MIC breakpoints recommended by the CLSI in 2015.⁸

Multiplex PCR for detection of genes conferring ESBLs, AmpC β-lactamases and carbapenemases was performed for all isolates with ertapenem MIC >0.5 mg/L. Whole genomic DNA of the isolates was extracted using a QIAamp DNA minikit and QIAcube instrument (Qiagen, Valencia, CA, USA) from colonies grown overnight on blood agar plate (Remel, Lenexa, KS, USA). Specific primers to detect ESBL alleles (bla_CTX-M, bla_TEM, bla_SHV, bla_VEB, bla_GES and bla_PER), plasmid-mediated AmpC genes (bla_ACC, bla_CMY, bla_MOX, bla_FOX, bla_DHAbla_ACT and bla_MIR) and bla_SPM, bla_GIM, bla_KPC, bla_VIM, bla_NDM, bla_IMP and bla_OXA genes were used as previously described.⁹

Statistical analysis

The χ² test was used to evaluate differences in percentages of isolate numbers between two indicated subgroups. Additionally, linear regression analysis was used to analyse the trend of annual rates on various β-lactamases among Enterobacteriaceae isolates. A P value <0.05 was considered statistically significant; all tests were two-tailed. All analyses were performed with the statistical package SPSS for Windows (version 17.0, SPSS, Chicago, IL, USA).

Results

Isolates harbouring various resistance alleles

During the study period, 2893 isolates with ertapenem MIC >0.5 mg/L were collected. Among these isolates, 2728 harboured β-lactamase genes. The β-lactamase-producing isolates comprised Escherichia coli (n = 1739, 63.7%), K. pneumoniae (n = 714, 26.2%), Enterobacter cloacae (n = 200, 7.3%), Proteus mirabilis (n = 21, 0.8%) and other species. Among the countries submitting >200 isolates, only IAI isolates collected from the Philippines and Vietnam had higher non-susceptible rates to imipenem (17.7% and 10.2%, respectively) than others (average, 7.9%).

Trends of antimicrobial susceptibilities and β-lactamase production

Among the 2728 β-lactamase-producing Enterobacteriaceae isolates, a stepwise and significant increase in annual rates of carbapenemase production (R = 0.793, P = 0.03), rising from 0.07% in 2008 to 1.1% in 2013, was found. Moreover, during 2008–10, IMP-26 enzyme was the predominant carbapenemase. However, the bla_NDM-1 allele became the most prevalent carbapenemase-encoding gene among the carbapenemase-producing isolates from 2011 onwards. For non-susceptible rates of antimicrobials evaluated against these IAI isolates, imipenem and amikacin showed the most favourable in vitro activities (8.0% and 9.3%, respectively), followed by ertapenem (20.7%), piperacillin/tazobactam (31.7%) and cefoxitin (36.7%), whereas prominently high non-susceptibilities were noted for levofloxacin (64.0%), ciprofloxacin (72.9%), cefepime (86.3%) and ceftriaxone (99.7%).

Proportions of CA-IAI versus HA-IAI isolates with different ESBL status

There was a significant difference in proportions of CA isolates between ertapenem-susceptible and ertapenem-non-susceptible subgroups of ESBL-positive E. coli isolates (31.7% versus 46.0%, P = 0.011), whereas no such difference was noted for ESBL-producing K. pneumoniae isolates (ertapenem-susceptible versus ertapenem-non-susceptible isolates, 30.9% versus 35.6%, P = 0.307).

Data on important ESBL and AmpC enzymes among IAI isolates

Table 1 shows detailed data on major ESBL enzymes and AmpC β-lactamases among 2673 isolates of important Enterobacteriaceae species. CTX-M-15, CTX-M-27 and CTX-M-14 variants were dominant ESBLs in E. coli isolates, while CTX-M-15 was the dominant variant in isolates of K. pneumoniae. In isolates of Enterobacter spp., AmpC enzymes (ACT and MIR) and carbapenemases played more important roles than ESBLs in conferring resistance to cephalosporins and ertapenem. Of AmpC enzymes, CMY-2 type (CMY-2) was the dominant variant in isolates of E. coli [87.7% (178/203)], the DHA-1 variant was the most common among isolates of K. pneumoniae [85.2% (109/128)], and ACT [86.1% (180/209)] and MIR [28.7% (60/209)] were the dominant variants among Enterobacter spp. isolates.

The MIC₉₀ values of ertapenem/imipenem for single CMY-2- and DHA-1-producing isolates were 4.0/8.0 and 4.0/2.0 mg/L, respectively. In addition, after excluding 11 isolates with undetermined in-hospital acquisition times, we found that 92.9% (13/14) of single CMY-1-producing isolates, 62.0% (116/187) of single CMY-2-producing isolates, 57.4% (70/122) of single DHA-1-producing isolates, 62.6% (87/139) of single ACT-producing isolates and 82.1% (23/28) of single MIR-producing isolates were HA in origin. Geographically, it is noteworthy that CMY-2 was the dominant variant among AmpC-positive E. coli isolates in most Asia-Pacific countries except South Korea and the Philippines. Additionally, the DHA-1 variant was less frequently seen among AmpC-positive K. pneumoniae isolates collected from South Korea than other Asia-Pacific countries.

Co-production of various β-lactamases among CA versus HA AmpC-producing isolates

Among 543 AmpC-positive Enterobacteriaceae isolates with clear in-hospital acquisition times, the isolates of CA-AmpC producers accounted for 35.9% of overall AmpC-producing ones. Of note, CA E. cloacae isolates were more likely to co-produce a carbapenemase [mostly NDM-1 (78%)] than E. cloacae isolates from HA infections (28% versus 9%, P = 0.0006).

Geographical distribution of various carbapenemases among IAI isolates

As shown in Table 2, no carbapenemase-producing Enterobacteriaceae isolates were identified in Hong Kong, South Korea, Kazakhstan, Singapore and New Zealand during this period. Of special note, some IAI bla_KPC-2-, bla_NDM-1-, bla_NDM-4- and bla_NDM-5-harbouring bacteria (including E. coli, K. pneumoniae, E. cloacae and Citrobacter freundii) were isolated in Vietnam during 2010–14. Likewise, many isolates of KPC-2-, NDM-1- and NDM-7-producing Enterobacteriaceae species have been found in the Philippines since 2011. Furthermore, one IAI bla_NDM-1-harbouring K. pneumoniae isolate was identified in Malaysia in 2013, as well as one in Australia and two in Thailand in 2014. Moreover, 10 bla_OXA-48-harbouring isolates, including E. coli (n = 3), K. pneumoniae (n = 5) and E. cloacae (n = 2), were identified in Vietnam from 2011 to 2014.

Discussion

In this survey, for proven β-lactamase-producing isolates, more than 30% of ESBL-positive isolates of E. coli and K. pneumoniae were obtained from patients with CA-IAIs. Similarly, a significant percentage (36%) of AmpC producers were also CA in origin. In addition, most of the E. cloacae isolates that co-produced carbapenemases (mainly NDM-1) and AmpC were acquired from community settings as well. These findings underscore that Asia is an epicentre of horizontal transfers of high-level resistance alleles between Gram-negative bacteria irrespective of community or nosocomial settings.

The plasmids of E. coli are prime disseminators of resistance genes because they are major reservoirs of ISEcp1 and ISCR1, important elements in mobilizing the bla_CTX-M-like, bla_CMY-2-like, bla_CMY-1-like and bla_DHA-1-like genes.^3,10 Compared with the IAI survey by Sheng et al.,⁷ CTX-M-14 and CTX-M-15 alleles, prevalent among E. coli and K. pneumoniae in Japan,¹¹ have been detected among a broader range of species (E. cloacae, Klebsiella oxytoca, Proteus mirabilis) across many Asian countries in our survey. Additionally, the high prevalence of CTX-M-27 among E. coli isolates could be attributed to those submitted by Vietnam (57.1%), consistent with one recent analysis.¹²

In contrast to our findings, several studies reported that there were great diversities of AmpC variants among E. coli isolates collected in some Asian countries.¹³ In our study, we found that 11.7% (203/1739) of all IAI E. coli isolates produced plasmid-mediated AmpC enzymes (mainly CMY-2). The rate was similar to that reported by Oteo et al.¹⁴ Moreover, we also found that the rate of plasmid-mediated AmpC production (17.9%, predominantly DHA-1) was higher among K. pneumoniae than among E. coli isolates (P < 0.001), resembling the rates reported by Yoo et al.¹⁵

Similar to bla_CTX-M and many AmpC-encoding alleles, ISs or transposons (ISKpn21, ISAba125; Tn4401; Tn1999 or Tn1999.2) also increase the mobilization potential of the bla_NDM-1, bla_KPC-2 and bla_OXA-48 alleles in some Enterobacteriaceae species.^16,17 In this study, we found that bla_NDM-positive Enterobacteriaceae species (primarily E. coli, K. pneumoniae and E. cloacae) were predominantly isolated from patients in Vietnam and the Philippines. Gram-negative bacteria harbouring bla_NDM alleles in clinical samples were first isolated in Vietnam in 2009¹⁸ and have been identified since 2010 in our study. Owing to the close cultural and economic links between Vietnam and India, it is not surprising that extensive person-to-person exchanges likely resulted in the emergence of bla_NDM-1-harbouring K. pneumoniae strains in Vietnam.¹⁹ In contrast, no other surveys address clinical bla_KPC-2- and bla_NDM-1,7-bearing E. coli, K. pneumoniae or E. cloacae isolates cultured from patients with IAIs hospitalized in the Philippines until now. Similarly, this is the first known study reporting detection of bla_OXA-48-harbouring Enterobacteriaceae isolates from IAI patients in Vietnam.

This study has some limitations. Firstly, as described by Rodríguez-Baño et al.,²⁰ the antimicrobial susceptibilities of CA isolates might be complicated by some healthcare factors, not investigated in this study. Secondly, high ertapenem MICs among Enterobacteriaceae spp. might be due to disappearance of important membrane porins and/or development of efflux pumps, which were not characterized in this investigation. Finally, the genetic relatedness and clones of special sequence types (e.g. ST131 E. coli, ST11 K. pneumoniae) of the enrolled isolates were not determined in this study.

In summary, the ESBL variants of CTX-M-14 and CTX-M-15 were prevalent in major Enterobacteriaceae species across all Asia-Pacific countries. The high CA rates among bla_CMY-2-harbouring E. coli and bla_DHA-1-harbouring K. pneumoniae isolates (both ≥38% CA) indicate that these two subsets along with Enterobacter spp. isolates pose enormous challenges to future IAI therapy. In addition, the IAI Enterobacteriaceae isolates collected from some countries (such as Vietnam and the Philippines) showed high levels of antimicrobial resistance. Consequently, there is an imminent need for improving antimicrobial prescription and infection control policies to halt the worsening trend in antimicrobial resistance among Enterobacteriaceae species causing IAIs.

Funding

This study was supported by Merck Sharp & Dohme.

Transparency declarations

None to declare.

Acknowledgements

We thank all of the investigators in the Asia-Pacific region for their participation in the SMART programme

Other investigators from the SMART Asia-Pacific Group

Tony Korman (Monash Medical Center, Clayton, VIC, Australia), Justin Ellem (Westmead Hospital, Westmead, NSW, Australia), Narelle George (Royal Brisbane Hospital, Brisbane, QLD, Australia), Geoffrey Coombs (Royal Perth Hospital, Perth, WA, Australia), Thomas Ling (Prince of Wales Hospital, Shatin, New Territories, Hong Kong), Owen Tsang (Princess Margaret Hospital, Hong Kong), V. Balaji (Christian Medical College, Vellore, India), Hiroshige Mikamo (Aichi Medical University Hospital, Nagakute, Japan), Shinya Kusachi (Toho University, Tokyo, Japan), Tetsu Mizutani (Osaka Police Hospital, Osaka City, Japan), Min-Ja Kim (Korea University Anam Hospital, Seoul, South Korea), In-Gyu Bae (Gyeongsang National University Hospital, Jinju, South Korea), Nurulhuda Binti Umur (Hospital Kuala Lumpur, Kuala Lumpur, Malaysia), Datin Ganeswrie Rajasekaram (Hospital Sul-tanah Aminah Johin Bahru, Johor Bahru, Malaysia), Susan Taylor (Middlemore Hospital at Counties Manukau District, Otahuhu, New Zealand), Sally Roberts (Auckland City Hospital, Grafton, New Zealand), Koen van der Werff (Wellington Hospital, Wellington, New Zealand), Dragana Drinkovic (North Shore Hospital, Auckland, New Zealand), Evelina Lagamayo (St Luke's Medical Center, Quezon City, The Philippines), Myrna Mendoza (Philippine General Hospital, Manila, The Philippines), Thean Yen Tan (Changi General Hospital, Singapore), Prabha Krishnan (Tan Tock Seng Hospital, Singapore), Ellie Wang (National Cheng Kung University Hospital, Tainan, Taiwan), Po-Liang Lu (Kaohsiung Medical University Hospital, Kaohsiung City, Taiwan), Chun-Eng Liu (Changhua Christian Hospital, Changhua City, Taiwan), Kenneth Yin-Ching Chuang (Chi-Mei Medical Center, Tainan City, Taiwan), Kwok-Woon Yu (Taipei Veterans General Hospital, Taipei, Taiwan), Yao-Shen Chen (Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan), Min-Chi Lu (Chung Shan Medical University Hospital, Taichung City, Taiwan), Siripen Panthuwong (Songklanakarin Hospital, Songkhla Province, Thailand), Pattarachai Kiratisin (Siriraj Hospital, Bangkok-Noi, Thailand), Nguyen Tran My Phoung (BinhDan Hospital, Ho Chi Minh City, Vietnam), Doan Mai Phuong (BachMai Hospital, Hanoi, Vietnam), Nguyen Thi Van (Benh Vien Viet Duc Hospital, Hanoi, Vietnam) and Tran Thi Thanh Nga (Cho Ray Hospital, Ho Chi Minh City, Vietnam).

References

Author notes