Prevalence of Carriage of Extended-Spectrum Beta-Lactamase-Producing Enterobacteriaceae Among Pregnant Women in the Primary Health Care Center and Hospital Setting in Indonesia

Background: The incidence of healthy individuals carrying Extended-Spectrum Beta-Lactamase producing Enterobacteriaceae (ESBL-E), especially E. coli (ESBL-EC) and Klebsiella pneumoniae (ESBL-KP), is increasing worldwide. ESBL-E causes early or late onset of neonatal sepsis, resulting in increased morbidity and mortality rates. Although maternal-neonatal transmissions of ESBL-E have been reported in several countries, the prevalence of ESBL-E carriage among pregnant women in Indonesia is not clear. In the present study, we compared the prevalence of carriage of ESBL-E among pregnant women in a primary health care center (PHC) versus two hospitals in Indonesia and identied the phenotypic and genotypic characteristics of the isolated ESBL-E strains. Methods: We collected rectal swab samples from 200 pregnant women who visited a PHC (101 women) or were admitted to Dr. Soetomo Referral Hospital or Airlangga University Hospital (99 women) in Surabaya, Indonesia from July to October 2018. The samples were cultivated on MacConkey Agar plates supplemented with cefotaxime 2 mg/L, at 37 o C overnight. The isolated strains were identied by Bruker MALDI Biotyper Ò System, phenotypic detection of ESBL was performed by the combination disk method, and antibiotic susceptibility was tested by the disk diffusion method. In addition, ESBL gene was identied by PCR and DNA sequencing and molecular epidemiological studies were performed by PFGE. Results: ESBL-E strains were isolated from 25 (rate of fecal carriage; 24.8%) pregnant women who visited the PHC and 49 (49.5%) pregnant women who were admitted to the hospitals. The rate of ESBL-E carriage of pregnant women in the hospitals was signicantly higher than that in the PHC. Among the 74 isolated ESBL-E strains, ESBL-EC was most frequently isolated (62 strains), followed by ESBL-KP (12 strains). In addition, bla CTX-M15 was the most frequent ESBL gene type of the isolated ESBL-E strains. Conclusion: Our results revealed the high prevalence of

in several countries (13,15). In Indonesia, despite the high prevalence of nosocomial and community ESBL-E carriage, the prevalence among pregnant women is not clear.
Since the 2000's, the community prevalence of ESBL-E is increasing globally (4,(16)(17)(18)(19). Surprisingly, a systemic review and meta-analysis of the prevalence of ESBL-E carriage among pregnant and post-partum women in Africa revealed that the prevalence of maternal carriage of ESBL-E in the community exceeded that in hospitals (22% versus 14%) (20,21). In the present study, we investigated the prevalence of ESBL-E carriage among 200 pregnant women who visited a primary health care center (PHC, 101 women) or were admitted to the hospitals (99 women) in Surabaya, Indonesia from July to October 2018, and identi ed the phenotypic and genotypic characteristics of the isolated ESBL-E strains.

Study design and setting
This cross-sectional observational study was carried out from July to October 2018 in Surabaya, Indonesia (22). The subjects of study consisted of 101 pregnant women who visited the Jagir Primary Health Center (PHC) and had not been hospitalized within a year, and 99 pregnant women who were hospitalized at Dr. Soetomo Hospital and Airlangga University Hospital for more than two days. A history of antibacterial medication was collected by a questionnaire.
Written informed consents were obtained from all participants, and the study was approved by the Ethical Committees from the Dr. Soetomo Hospital (No. 0353/KEPK/VI/2018) and Airlangga University Hospital (No. 193/KEH/2018).

Isolation and identi cation of bacterial strains
Rectal swab samples were taken using Amies transport media and the samples were immediately cultured on MacConkey selective media supplemented with 2 µg/ml cefotaxime at 37 o C, overnight. Suspected single colonies were sub-cultured in blood agar at 37 o C, overnight. The bacterial isolates were identi ed using a MALDI Biotyper (Bruker Daltonics K.K., Yokohama, Japan).

Antimicrobial Susceptibility Testing and ESBL Screening
A Kirby-Bauer disc diffusion test was performed using Muller-Hinton agar plates for the antimicrobial susceptibility test according to the Clinical and Laboratory Standards Institute (CLSI) recommendations (23) for the following 16 antibiotics: ceftazidime (CAZ), cefotaxime (CTX), cefepime (FEP), ampicillin (AMP), imipenem (IPM), meropenem (MEPM), amikacin (AMK), minocycline (MI), tetracycline (TE), Trimethoprim-Sulfamethoxazole/co-trimoxazole (ST), nalidixic acid (NA), cipro oxacine (CPFX), cefmetazole (CMZ), omoxef (FMOX), aztreonam (AZT), and piperacillintazobactam (TZP) (24). The susceptibilities for colistin (CL) and tigecycline (TGC) were determined by E-test (bioMerieux, Marcy-l'Étoile, France). The European Committee on Antimicrobial Susceptibility Testing (EUCAST) criteria were used for CL and TGC of Enterobacteriacea (25). ESBL screening was initially performed with the CLSI con rmatory test, using both cefotaxime (CTX/30 mg) and ceftazidime (CAZ/30 mg) disks alone and in combination with clavulanic acid (CA/10 mg) (Eiken Chemical, Tokyo, Japan). The test was considered positive when the diameter of the growth-inhibitory zone around either the CTX or the CAZ disk in combination with CA increased by ≥ 5 mm compared to the growthinhibitory zone around the disk containing CTX or CAZ alone (23).  Table 1, as previously described (9). The PCR positive samples were further examined for DNA sequencing to identify the type of bla-gene. The DNA sequencing of the PCR products was carried out by Euro ns Genomics (Tokyo, Japan) after puri cation of the PCR products by the QIAquick PCR Puri cation Kit (Qiagen, Hilden, Germany). The DNA alignments of ESBL-encoding genes were examined by the BLAST program (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The isolates of ESBL-EC were classi ed into phylogenetic group A, B1, B2 or D by using multiple PCR detecting chuA, yjaA and TspE4.C2 as previously described (9,26,27).

Pulsed-Field Gel Electrophoresis (PFGE)
PFGE analysis was performed for the ESBL-EC harbouring the bla CTX−M−15 gene. Preparation of total genomic DNA for PFGE was performed as follows: One loop (1 µL) colonies from an overnight culture were suspended in 150 µL TE buffer (10 mmol/liter Tris-HCl, pH 8.0, 1 mmol/liter EDTA). Chromosomal DNA was prepared in solid agarose plugs by mixing 150 µL bacterial cell suspension with an equal volume of 1% low-melting-point agarose (SeaKem® Gold Agarose, Lonza Rockland, Inc., Rockland, ME) and then incubated overnight at 50 °C in lysis buffer (50 mmol/liter Tris-HCl, pH 8.0, 50 mmol/liter EDTA, 1% laurylsarcosine, 1 mg proteinase K/ml). Thin slices of plug were washed three times for 20 min in sterile distilled water and twice for 20 min in TE buffer. The plugs were digested overnight with 30 U of XbaI (TaKaRa, Shiga, Japan) restriction enzyme according to the manufacturer's instructions (28). PFGE was performed with the GenePath System (Bio-Rad Laboratories Inc., Hercules, Calif., USA) using 1% SeaKem® Gold Agarose (Lonza Rockland, Inc., Rockland, ME) in 2 liters of 0.5X Tris-buffered EDTA running buffer. The electrophoretic conditions were as follows: initial switch time, 5.3 s; nal switch time, 49.9 s; run time, 20 h; voltage, 6 V/cm; buffer temperature, 14 °C. Lambda Ladders (Promega, Madison, WI, USA) were used as molecular weight standards. A dendrogram was generated using GelJ ver2 (29) calculated using the unweighted pair group average (UPGMA) and the Jaccard coe cients. Strains were de ned as having a clonal relationship if they possessed a similarity higher than 90% to the PFGE pro les (30).

ESBL-E carriage of pregnant women in hospitals versus PHC
In this study, rectal swab samples were collected from a total of 200 pregnant women (PHC: 101; hospital, 99) from July to October 2018 in Surabaya, Indonesia. Table 2 summarizes the characteristics and ESBL-E carriage rates in the study population. The rates for pregnant women who were older than 35 years and had a history of antibiotics use were signi cantly higher in the hospital group than in the PHC group (p < 0.01). ESBL-E strains were isolated from 25 (rate of faecal carriage; 24.8%) pregnant women in the PHC group and 49 (49.5%) pregnant women in the hospital group. ESBL-EC strains were isolated from 20 (19.8%) pregnant women in PHC and 42 (42.4%) pregnant women in hospital. The rates of ESBL-E and ESBL-EC carriage of pregnant women in hospital were signi cantly higher than those in the PHC (p < 0.01). Among the 74 isolated ESBL-E strains, ESBL-EC was most frequently isolated (62 strains, carriage rate: 31.0%), followed by ESBL-KP (12 strains, carriage rate: 6.0%).

Antimicrobial susceptibility of ESBL-E isolates
Antimicrobial susceptibility testing was performed for a total of 74 isolates of ESBL-E. The rates of drug-resistant strains are shown in Fig. 1. All isolates tested were resistant to AMP, and sensitive to carbapenems (IPM and MEPM), AMK and TGC. Only ESBL-EC and KP in the hospitals showed resistance to MI and TZP (EC 27.9% and KP 14.3% to MI, EC 27.9% and KP 42.9% to TZP). Among all strains tested, only one ESBL-EC isolate in hospital was resistant to CL. As for the susceptibilities of ESBL-EC strains against TE, ST, NA, CPFX and AZT, higher numbers of hospital strains showed resistance than PHC strains, whereas higher numbers of ESBL-EC in PHC strains showed resistance to FMOX and CMZ than hospital strains.

Identi cation of ESBL gene type
In the 62 isolates of ESBL-EC and 12 isolates of ESBL-KP, bla CTX−M15 was the most frequently identi ed ESBL gene type in 19 (30.6%) of ESBL-EC isolates and 6 (50%) of ESBL-KP isolates (Table 3). Furthermore, the bla CTX−M−15 gene was detected in 11 (55%) strains of ESBL-EC in PHC. This rate was signi cantly higher than the rate of 19% ESBL-EC in the hospitals (p < 0.01). As well as the bla CTX genes were also identi ed in this study (Table 3).

Discussion
Since the rst community-acquired ESBL-E was reported in the late 1990s, the community carriage rate of ESBL-E has continuously increased worldwide (16,31,32). Reported ESBL-E community carriage rates have increased more rapidly in some regions such as Southeast Asia, Eastern Mediterranean, and Western Paci c during the 2000s compared to other regions. Recent studies from these regions often reported an ESBL-E community carriage rate around 50% (16). In the present study, ESBL-E strains were isolated from 74 out of 200 (37%) pregnant women in PHC and hospital settings ( Table 2) The use of antibiotics is considered as the primary cause of the spread of antimicrobial resistance, and it has been reported that global consumption of antibiotics increased by 65% from 2000 to 2015. At present, there is a strict demand for proper use of antibiotics in the medical eld, and the consumption of antibiotics in developed countries tends to be suppressed. In emerging countries, it continues to increase rapidly (34). In the present study ESBL-E strains were more frequently isolated from pregnant women in hospitals (49.5%) than pregnant women in the PHC setting (24.8%) (p < 0.01, Table 2). Consistent with this nding, the rate of antibiotics use in pregnant women in hospitals (41.4%) was signi cantly higher rather than the pregnant women in PHC (19.8%) (p < 0.01, Table 2). Antimicrobial susceptibility testing showed that all 74 ESBL-E strains were sensitive to carbapenems, including IPM, MEPM, and also sensitive to AMK and TGC. However, ESBL-EC strains isolated from hospitals were more resistant to MI, TE, NA, CL, ST, AZT and TZP compared to ESBL-EC strains isolated from the PHC (Fig. 1). On the other hand, ESBL-EC strains isolated from the PHC showed higher resistance compared to hospitals against FMOX and CMZ, but these resistance rates were still comparatively low (Fig. 1).
When the emergence of ESBL-E strains began in hospitals in the 1980s, the strains were mostly Klebsiella spp. and Enterobacter spp. which contained bla TEM or bla SHV genes (35,36). In contrast, community-acquired ESBL-E strains emerging after the 1900s were mostly E. coli harboring the bla CTX−M ESBL gene (37). In the present study isolated ESBL-E strains were mostly ESBL-EC (62 out of 74 ESBL-E strains), and 32 of the 62 ESBL-EC strains harbored the bla CTX−M ESBL gene (Table 3) (Table 3). In addition, the bla CTX−M−15 genotype in ESBL-EC was more frequently identi ed in pregnant women in PHC (55%) than in hospitals (19%). Consistent with these ndings, the literature review by Bevan et. al. (38) reported that CTX-M15 has been the most common ESBL genotype in Southeast Asia as well as most other parts of the world in the past two decades. The review also mentioned the decline of CTX-M2 and the emergence of CTX-M27, which is a single-nucleotide variant of CTX-M14 (38). In the present study, we detected the bla CTX−M27 gene in 2 strains of ESBL-EC and one strain of ESBL-KP, but we detected no bla CTX−M2 in any ESBL-E strains.

Conclusion
In conclusion, we investigated the prevalence of ESBL-E carriage in pregnant women in the PHC and hospital settings.
Our results revealed a high carriage rate of ESBL-EC in pregnant women (24.8% in PHC and 49.5% in hospitals) in Surabaya, Indonesia. The ESBL-E carriage rate was signi cantly higher in pregnant women in hospitals than in the PHC, along with a signi cantly higher rate of antibiotics use (19.8% in the PHC and 41.4% in hospitals). The isolated ESBL-E strains were ESBL-EC (62 strains) and ESBL-KP (12 strains), and bla CTX−M−15 was the most prevalent (30.6%) ESBL gene type. The clonal relatedness between PHC and hospital isolates was not found in this study. Continuous surveillance for ESBL-E carriage in pregnant women is strongly recommended to reduce the incidence of neonatal sepsis in Indonesia.  . In addition, all participants signed written informed consent for participation in the study.

Abbreviations
Consent for publicationThis manuscript does not contain any individual person's data in any form.
Competing interestsNone to declare.  Tables   Table 1. Primer sets for PCR detection of ESBL-encoding genes   Target Primer name Sequence (