Risk factors and molecular features of sequence type (ST) 131 extended-Spectrum-β-lactamase-producing Escherichia coli in community-onset female genital tract infections

Escherichia coli (E. coli) is known to cause urinary tract infection (UTI), surgical site infection, and meningitis in neonates, as well as existing as a commensal flora of the human gut [1, 2]. The colonization of E.coli has been reported in both pregnant (24–31%) and non-pregnant (9–28%) women [3, 4]. The clinical risk factors for E. coli colonization or infection in the female genital tract is not fully understood, but postmenopausal changes including cessation of estrogen production, increased vaginal pH, disappearance of lactobacilli with the concurrent colonization of Enterobacteriaceae including E. coli is thought to play a role.⁵ Pregnant women with E. coli infection prompt additional attention due to its consequences regarding the pregnancy outcome including the infection of the new born [5].

In the past decade, colonization or infection of extended-spectrum-β-lactamase (ESBL)-producing E. coli has remarkably increased in the community, mostly due to the spread of high virulent and multidrug-resistant sequence type 131 (ST131) E. coli clone [6‐8]. According to recent epidemiologic studies based on community-onset UTI, the situation of Korea is not different [6‐8]. The presence of ESBL-producing E. coli in female genital tract infections are scarcely identified within our knowledge.

For the prospective observational study, 91 non-duplicated E. coli isolates from the female genital tract specimens of 514 consecutive, sequentially encountered patients were analyzed. All specimen were retrieved from one community hospital in Gyeonggi-do province, South Korea (742 beds) under the diagnosis of community-onset vaginitis or cervicitis between June 2016 and April 2017. All patients had infection signs such as increased leucorrhoea, vaginal dyspareunia, intermittent pruritus, burning sensation, and foul odor. Sites of the acquisition were determined as described by Friedman with some modifications [9]. Community-onset was defined as diagnosis given within 48 h of admission, further categorized as community-onset healthcare-associated (COHA) and community-associated (CA) group. COHA infections had any one of the following histories: attended a hospital or hemodialysis clinic or received intravenous chemotherapy in the 30 days before the infection; hospitalized in an acute care hospital for 2 or more days in the 90 days, transfer-in from other healthcare facility before the infection. Others were defined as CA infection [9].

Species Identification and susceptibility testing were performed with Microscan Walk-away plus system (BeckmanCoulter, Inc., CA, USA) and MicroScan Neg Breakpoint Combo Type 44 (Siemens Healthcare Diagnostics Inc., West Sacramento, CA, USA). Antimicrobial susceptibility of the 91 E. coli isolates was tested and interpreted using CLSI criteria [10]. ESBL production was confirmed by ESBL double-disk synergy test [11], and ESBL genotype was determined by PCR and sequencing [12]. For the detection of ST131, all isolates were screened by PCR for O16-ST131, and O25b-ST131 [13]. Pulsed-field gel electrophoresis (PFGE) was performed as described in our previous study [14]. The patterns were analyzed using InfoQuest FP software (Bio-Rad) to generate a dendrogram based on the unweighted pair group method, with an arithmetic average (UPGMA) from the Dice coefficient with 1% band position tolerance and 0.5% optimization settings. A PCR based replicon typing (PBRT) were schemed in the 18 ESBL-producing E. coli isolates, targeting the replicons of the major plasmid families occurring in Enterobacteriaceae (HI2, HI1, I1-γ, X, L/M, N, FIA, FIB, FIC, W, Y, P, A/C, T, K, B/O) according to the protocol by Carattoli, et al. [15] Ten of the 18 isolates were successfully conjugated and PBRT was performed in these 10 transconjugants.

Statistical analysis was performed using Chi-square test for the comparative analysis of categorical variables to determine the independent risk factors. Fisher’s exact test was used when 25% of the cells had an expected frequency of less than 5. Odds ratio (OR) and 95% confidence interval (CI) values were calculated for binomial variables. Variables were first compared using univariate logistic regression, then a multivariate analysis using a backward selection with variables with a p-value < 0.1 in the univariate study was carried out. Multivariate logistic regression model was adjusted for age, pregnancy, intrauterine device use, menopause, admission from the nursing home, PID, recurrent vaginitis, UTI and previous antibiotic treatment within one month. Statistical significance was defined as p < 0.05. SPSS 17.0 (SPSS, Chicago, IL, USA) and SAS 9.4(SAS Institute Inc., Cary, NC) was used.

Of the 514 consecutive specimens obtained from the infected female genital tract, 17.7% (91/514) had E. coli infection, of which 19.8% (18/91) were ESBL producers. All ESBL-producers had the CTX-M genotypes; CTX-M-15 was the most common type (n = 15), one of which also had CTX-M-27. CTX-M-55 producers were also detected (n = 3). Of the total 19 E. coli isolates, O25b-ST131 accounted for 15.4% (14/91) and O16-ST131 clones for 6.6% (6/91). Seven of the 18 ESBL producers were a ST131 clone with four being O25b-ST131 and three O16-ST131 (Table 1 and Additional file 1: Table S1). PFGE patterns of ESBL-producing E. coli showed three dominant clonal groups with the cut off 80% similarity, suggesting the clonal spread of ESBL-producing E. coli in the community (Fig. 1). In plasmid analysis, ten isolates succeeded in conjugation and plasmid types were IncFII (n = 4), IncFI (n = 3), IncI1-Iγ (n = 3) and one non-typable (Table 1). Antimicrobial susceptibility of the 91 E. coli isolates were analyzed. ST131 clone showed high resistance rates (RRs) to both 3rd generation cephalosporin and fluoroquinolones; 36% (O25b-ST131) versus 20% (non-ST131) to cefotaxime and 50% (O25b-ST131) versus 35% (non-ST131) to ciprofloxacin (Table 2 and Additional file 2: Table S2).

The median age of the patients with ESBL-producing E. coli was 45 years old (31–96) and accompanying diseases were recurrent vaginitis (n = 4), atrophic vaginitis (n = 3), pelvic inflammatory disease (n = 3), preterm labor (n = 3), and UTI (n = 2). They were treated with a variety of antimicrobial therapies, which included cephalosporin, carbapenems, piperacillin-tazobactam, and amikacin, alone or in combination and achieved a clinical cure. All three pregnancies with ESBL-producing E. coli female genital tract infection were prematurely terminated due to preterm labor (Table 1).

Independent clinical risk factors for acquiring the ESBL-producing E. coli over non-ESBL producing E. coli were preterm labor (OR, 10.911, 95% CI, 1.199–99.301; P = 0.0339), intrauterine device insertion (OR, 6.460; 95% CI, 1.004–41.568; P = 0.0495), history of recurrent vaginitis within one month (OR, 40.130; 95% CI, 9.980–161.366; P ≤ 0.0001), history of UTI within one month (OR, 18.915; 95% CI, 5.469–65.411; P ≤ 0.0001), and antibiotics treatment within one month (OR, 68.390; 95% CI, 14.870–314.531; P ≤ 0.0001) (Table 3) .

Clinical characteristics were compared between the ST131 and non-ST131 clone. In the multivariate logistic regression analysis adjusting for the compounding variables, independent risk factors with statistical significance could not be found that were associated with acquiring the ST131 clone (Table 3).

ESBL-producing E. coli was considered to be a crucial nosocomial pathogen when it was first isolated in the late 1980s [6‐8]. With the occurrence of the ST131 clone, ESBL-producing E. coli in the community setting has been widely noted in recent studies [16‐18]. Our previous study showed that 27% (58/213) of E. coli isolates from UTI patients belonged to the globally epidemic ST131 clone [6]. The Korean Antimicrobial Resistance Monitoring System (KARMS) reported the resistance rates of E. coli to cefotaxime to be 35% in 2015 [19]. Asymptomatic carriage of ESBL-producing E. coli among healthy individuals playing the role of reservoir are also known to have increased [20]. Fecal colonization with ESBL-producing E. coli are also on the rise [21].

Analysis of vaginal microbiome showed region-specific variation, setting the basis of the need for a region-specific data of the epidemic multi-drug resistant ST131 ESBL-producing E. coli clone [22, 23]. We evaluated the prevalence of ESBL-producing E. coli, focusing on the ST131 clone in female genital tract infections because it could be another indicator of the spread of highly virulent and multi-drug resistant E. coli in the community, which could be an emerging threat to the public health.

Of the 514 consecutive specimens obtained from the infected female genital tract, 17.7% (91/514) had E. coli infection, of which 19.8% (18/91) were ESBL producers. Fourteen of the 18 ESBL producers were community-associated(CA) without any history of hospitalization, suggesting the possibility of community based spread of ESBL-producing E. coli (Table 1).

CTX-M-15 was the most common type (n = 15) and this was in accordance of the previous isolates of CTX types in Korea in which either CTX-M-1 group (including CTX-M-15 or CTX-M-55) or CTX-M-9 group (including CTX-M-14 or CTX-M-27) prevailed [6, 7]. CTX-M-55 genotype is a variant of CTX-M-15 with a single amino acid substitution, which was known to frequent in China and considered as a potential threat to community spread [24]. Recently CTX-M-55-producing Shigella and Salmonella were isolated in Korea with bla_CTX-M-55 genes inserted into IncI1, IncA/C, and IncZ plasmid, downstream of ISEcp1, IS26-ISEcp1 and ISEcp-IS5 sequences, which suggests CTX-M-55 dissemination to different bacterial species by lateral plasmid transfer [25]. In this study, ten of 18 ESBL-producing E. coli succeeded in conjugation, and plasmid types were IncFII (n = 4), IncFI (n = 3), IncI1-Iγ (n = 3) with one non-typable case. Most of ESBL-producing E. coli showed similar clonality in PFGE, which suggests that dominant clone of CTX-M type ESBL-producing E. coli in conjugative plasmids are circulating in the community.

O25b-ST131 and O16-ST131 clones accounted for 15.4% (14/91) and 6.6% (6/91), respectively, suggesting a high prevalence of ST131 E. coli clone in the female genital tracts of Korean women. The expansion of the ST131 E. coli clone with phylogenetic B2, serotype O25b, fimH type H30 is suggested to reveal multidrug resistant property [8, 17]. In this study, O25b-ST131 clones showed high rates of multidrug-resistancy against the 3rd generation cephalosporins as well as fluoroquinolones (Table 2 and Additional file 2: Table S2).

Compared to the ESBL-nonproducing E. coli group, ESBL-producing E. coli group showed higher tendency of having clinical features such as history of recurrent vaginitis within 1 month (OR, 40.130; 95% CI, 9.980–161.366; P ≤ 0.0001), history of UTI within 1 month (OR, 18.915; 95% CI, 5.469–65.411; P ≤ 0.0001), and history or antibiotics treatment within 1 month (OR, 68.390; 95% CI, 14.870–314.531; P ≤ 0.0001). E. coli is a well-known etiologic agent for UTI, so it is not surprising that female genital tract infection with ESBL-producing E. coli were strongly associated with UTI [6‐8]. Previous antibiotic exposure might have resulted in the selection pressure of a resistant clone, concomitantly developing multidrug-resistancy [14, 26].

Aerobic vaginitis is a recently defined vaginitis type, which differs from the common bacterial vaginosis in terms of scarce presence of lactobacilli and positive cultured aerobic bacteria including group B streptococci, E.coli, and enterococci [5]. In postmenopausal women with atrophic vaginitis, the lack of estrogen results in deficiency of mucosal epithelial barrier, lactobacilli disappear from the vaginal flora decreasing the pH of the vagina, resulting in predominant colonization by Enterobacteriaceae, especially E. coli., serving an adequate environment for aerobic vaginitis to set place [2, 27].

In cases of pregnancy, aerobic vaginitis is known to be associated with an increased risk of preterm labor and chorioamnionitis [5]. Although all three pregnancies infected by ESBL producers resulted in preterm labor in this study, due to the small number of included patients and the omission of cases without genital tract infection, the role of ESBL producers in poor pregnancy outcome should be interpreted with caution. Further studies including a larger number of pregnancies to elucidate the role of ESBL-producing E. coli in the genital tract infection is warranted.

In conclusion, a dominant clone of CTX-M type ESBL-producing E. coli in conjugative plasmids seems to be circulating in the community and considerable number of ST131 E. coli clone in the genital tracts of Korean women was noted. Sustained monitoring of molecular epidemiology and an adequate control of the clinically high-risk group is needed to prevent ESBL-producing E. coli from spreading throughout the community.