In this study each participant was allowed to serve as her own control; thus, avoiding the probability of between-population differences that might lead to comparisons of UTI isolates with fecal isolates from healthy hosts. This is an exceptionally powerful method to remove the effects of known or unrecognized potential confounders, such as behavioral, environmental, physiological, or genetic differences between infected and uninfected hosts. In this study, we assessed 54 pregnant women with recurrent UTI in their first trimester. We compared E. coli isolated urine and fecal flora samples in terms of quinolone-resistant and ESBL, and we also investigated whether recurrence could be predicted by the characteristics of the E. coli fecal flora.
The results of our study indicate that the prevalence of antibiotic resistance in UPEC in comparison with fecal flora samples is close to each other. In both groups the most effective antibiotic against E. coli isolates was imipenem. And the high incidence of resistance was observed for nalidixic acid resistance (77.8% for UPECs and 76% for fecal flora isolates). These data are in agreement with the results of Dellgren et al.[16] Sweden and Bahadori1 et al. [11, 17] in Iran. Hence, nalidixic acid has been widely used to treat acute lower UTI, which has become nearly ineffective to treat UTI in our country. Also, more than 50% of the isolates were insensitive to amoxicillin–clavulanic acid and co-trimoxazole in both groups. This finding is in line with an earlier study in Iran[18].
According to the findings, the ESBL-producing phenotype were detected in 44.4% (24/54) of the urinary isolates, which was slightly lower than a fecal flora E. coli isolates 48.1%(26/54). The incidence of ESBL phenotype might vary across geographical regions with low rates of about 1.5% reported in Denmark[19] and 5% in Canada [20] compared to much higher prevalence rates documented in other countries[21, 22]. In the current study, the presence and identification of ESBLs genes were determined by multiplex PCR. Our statistics showed that blaTEM−1 was the most prevalent ESBL gene followed by bla CTX−M and blaSHV in both groups. In comparison to a similar study conducted in Kerman, the most frequent ESBL gene was reported to be bla CTX−M, blaTEM−1, and blaSHV genes, respectively[23]. Frequently, ESBL producers are resistant to other antibiotics, such as fluoroquinolones[6]. Based on the literature, fluoroquinolone resistance rate is ever increasing, even more than 50%, raising serious concerns in Iran and other parts of the world[24, 25]. In the present study, nearly half of the isolates were resistant to fluoroquinolones in UPEC and fecal flora samples (Table 2). In this study, higher resistance rate of strains against nalidixic acid and ciprofloxacin in UPEC and fecal flora samples are very close to the other findings of a recent research [26].
Table 5
Distribution of qnr genes in relation with quinolone resistance in fecal flora samples
Genes Antibiotic | Pattern | qnrS | P-value | qnrB | P-value | qnrA | P-value | aac(6′)- Ib-cr | P-value |
Positive No.(%) | Positive No.(%) | Positive No.(%) | Positive No.(%) |
Nalidixic acid | R:41 S:13 | 30 (73.1) 4 (30.7) | 0.45 | 13 (31.7) 2 (15.4) | 0.07 | 3 (7.3) 0 (0) | 0.47 | 6 (14.6) 1 (7.7) | 0.5 |
Ciprofloxacin | R:30 S:24 | 25 (83.3) 9 (37.5) | 0.45 | 11 (36.7) 4 (16.7) | 0.46 | 2 (6.7) 1 (4.2) | 0.6 | 7 (23.3) 0 (0) | 0.2 |
Ofloxacin | R:25 S:29 | 15 (57.7) 19 (67.9) | 0.54 | 9 (34.6) 6 (21.4) | 0.38 | 3 (12) 0 (0) | 0.5 | 6 (24) 1 (3.4) | 0.53 |
Levofloxacin | R:27 S:27 | 16 (59.3) 18 (66.7) | 0.42 | 10 (37) 5 (18.5) | 0.63 | 2 (7.4) 1 (3.7) | 0.7 | 5 (18.5) 2 (7.4) | 0.21 |
Norfloxacin | R: 28 S:26 | 17 (65.4) 17 (60.7) | 0.57 | 9 (32.1) 6 (23.1) | 0.2 | 2 (7.1) 1 (3.9) | 0.6 | 5 (17.9) 2 (7.7) | 0.53 |
Our results showed that resistance to the tested fluoroquinolones in ESBL-producing isolates was significantly higher than in non-ESBL-producing isolates. Although some studies reported that there was no significant association between resistance to fluoroquinolone and ESBL-producing isolates[27, 28], our results revealed that there was a significant relationship between resistance to nalidixic acid and ofloxacin in ESBL-producing isolates in fecal flora and resistance to nalidixic acid in ESBL-producing isolates in UPEC samples. Co-resistance to beta-lactams and fluoroquinolones can be related to the presence of ESBL and some of the quinolone-resistant genes in the same mobile genetic elements[29].
The qnrS (53.7% in urinary and 63% in fecal flora isolates) was the most prevalent PMQR gene in this study, which is in agreement with previous reports [30, 31]. Meanwhile, significant association was only observed between the presence of qnrS genes and nalidixic acid and ciprofloxacin resistance in fecal flora samples. In contrast, qnrA, qnrB and aac(6′)-Ib-cr were detected at low frequency. The rate of qnrB was more frequently was expressed in fecal flora (27.7%) in comparison with UPEC (15%) samples. It seems that the Qnr family can create a favorable condition for quinolone resistance. In addition, point mutations in the gyrA and parC genes might play a role as a principal mechanism of fluoroquinolones resistance [32]. Our data indicated that qnrS, qnrB and aac(6′)-Ib-cr gene was detected in significant proportion of the ESBL-producing in UPEC and fecal flora samples. However, no significant association was observed between the presence of qnr genes and ESBL-producing isolates, but it was detected only between aac(6′)-Ib-cr gene and ESBL-producing in UPEC samples (P = 0.034). Interestingly, at least 1 ESBL was detected in PMQR-positive isolates. Several previous studies reported a high prevalence of qnr genes among ESBL-producing isolates[33, 34].
Many studies have revealed that urine isolates collectively differed dramatically from fecal flora isolates, concerning antibiotic resistance content profiles, suggesting an increased resistance potential of the urine isolates[35]. In this study, comparing the results revealed a significant diversity in phenotypic drug sensitivity profiles as well as the distribution of ESBL-encoding genes and PMQR genes in urinary and fecal flora isolates. Nevertheless, phenotypic drug sensitivity profiles in 7 patients were similar to each other, and only four out of the 7 isolates had completely identical antibiotic resistance profiles and genotypes (Table 2).
In our search, a distinct difference between UPEC and commensal E. coli isolates was observed, with respect to their resistance pattern and distribution in qnr and ESBL genes. However, it is reasonable to say that UPEC and commensal E. coli isolates might have similar characteristics for adapting to an extraintestinal lifestyle, which in turn allows commensal E. coli to cause extraintestinal infections in humans as well as UPEC. As previously mentioned, commensal E. coli can serve as reservoirs of virulence and resistance genes for human pathogenesis.