Prospective Molecular and Phenotypic Characterization of blaCTX-M and blaTEM ESBL Genes in Uropathogenic Escherichia coli from ICU Patients at Three Tertiary-Care Hospitals in Tabriz, Iran

Abstract

Background

Urinary tract infections (UTIs) caused by extended-spectrum β-lactamase (ESBL)-producing Escherichia coli have become a major therapeutic challenge worldwide, especially in intensive care units. In northwestern Iran, the molecular epidemiology of the key resistance genes blaCTX-M and blaTEM, along with their potential co-carriage on plasmids, remains poorly understood. This gap limits the development of effective local empirical treatment guidelines and infection control strategies.

Methods

A cross-sectional study was conducted at three tertiary-care hospitals in Tabriz, Iran (2025). We identified 100 non-duplicate Escherichia coli isolates recovered from ICU patients with UTIs using standard microbiological methods. Antimicrobial susceptibility testing was performed by the disk diffusion method according to CLSI 2024 guidelines. ESBL phenotype was confirmed using the combined disk test with Ceftazidime and Clavulanic acid. Genotypic detection of blaTEM and blaCTX-M was performed by conventional PCR. Chi-square and Fisher’s exact tests were used to assess statistical associations.

Results

Among 135 culture-positive urine specimens, E. coli was the predominant pathogen (74.07%), followed by Klebsiella pneumoniae (11.85%). Universal resistance to Cefotaxime (100%) and high resistance to Ceftazidime (95%) were observed. Imipenem retained the highest in vitro activity (82% susceptibility). Phenotypic ESBL production was detected in 48/100 (48%) isolates. Among ESBL producers, blaTEM was present in 28 (58.3%), blaCTX-M in 20 (41.7%), and both genes co-occurred in 18 isolates (37.5%). The observed co-occurrence significantly exceeded expected values under independence (observed: 18 vs. expected: 11.7; Chi-square p = 0.0005; Fisher’s exact p = 0.00025), demonstrating non-random co-occurrence consistent with, but not proof of, plasmid-mediated co-carriage.

Conclusion

ESBL-producing E. coli carrying both blaTEM and blaCTX-M are highly prevalent among ICU-associated UTI in Tabriz. The statistically significant non-random co-occurrence of these genes has direct implications for surveillance strategies, necessitating simultaneous genotyping of both genes. Carbapenems remain effective, underscoring the urgency of their stewardship to prevent the emergence of carbapenem-resistant Enterobacterales (CRE) in this setting.

Introduction

Urinary tract infections (UTIs) are among the most common bacterial infections in both community and healthcare settings, with a global burden exceeding 400 million episodes annually.¹ Escherichia coli remains the predominant uropathogen, responsible for 75–85% of community-acquired and up to 65% of nosocomial UTIs. Uropathogenic E. coli (UPEC) strains possess a wide array of virulence factors, including type 1 and P fimbriae, α-hemolysin (HlyA), cytotoxic necrotizing factor-1, and multiple siderophore systems that facilitate adhesion, colonization, biofilm formation, and evasion of host defenses.^2,3

The clinical management of UTIs has become increasingly challenging due to the rapid rise of antibiotic-resistant strains, which have compromised the effectiveness of first- and second-line agents and driven greater reliance on carbapenems as last-resort therapies.⁴ Among the various resistance mechanisms, the dissemination of extended-spectrum β-lactamases (ESBLs) in Enterobacterales stands out as one of the most significant antimicrobial resistance challenges of the 21st century.⁵ ESBLs are class A serine β-lactamases capable of hydrolyzing third- and fourth-generation cephalosporins and monobactams, while remaining susceptible to β-lactamase inhibitors such as clavulanic acid.

Among the major ESBL families, TEM (Temoniera) and CTX-M (cefotaximase-Munich) enzymes predominate. Since the early 2000s, CTX-M-type enzymes have largely replaced TEM and SHV variants as the leading ESBLs worldwide—a shift often referred to as the “CTX-M pandemic”.^6,7 These resistance genes are typically located on self-transmissible conjugative plasmids that frequently carry multiple resistance determinants, facilitating rapid horizontal gene transfer (HGT) between bacterial strains and species. This process, known as co-transfer, significantly accelerates the spread of multidrug resistance.⁸

Despite the expanding global literature on ESBL epidemiology, important knowledge gaps remain, particularly in Iran. Most molecular epidemiological studies from Iran were conducted before 2015 and predate the COVID-19 pandemic, a period marked by a sharp increase in antibiotic consumption—especially β-lactams in hospital settings. This surge likely accelerated the dissemination of resistance genes.⁹ Furthermore, while the individual prevalence of blaTEM and blaCTX-M has been reported in various Iranian cities, the co-occurrence and potential co-transfer of these two genes on shared plasmids have not been systematically investigated in northwestern Iran, particularly in Tabriz, the region’s largest metropolitan area. Additionally, existing regional studies have largely focused on community-acquired isolates from outpatient settings, leaving ESBL-producing E. coli in intensive care unit (ICU)-associated UTIs—where antibiotic pressure is highest—poorly characterized.¹⁰

Understanding the co-occurrence of blaTEM and blaCTX-M is critical. If these genes are co-located on the same plasmid, surveillance and control strategies focused on a single gene may inadvertently promote the spread of the other. Plasmid-mediated co-transfer can also create stable reservoirs of multidrug resistance that spread rapidly through successful clonal lineages, such as the pandemic E. coli sequence type 131 (ST131).¹¹ From a clinical perspective, up-to-date molecular data from high-risk ICU environments are essential for guiding empirical therapy, updating local antibiograms, and implementing appropriate infection control measures.¹²

The present study was therefore designed to determine the prevalence of ESBL-producing E. coli in ICU-associated UTIs at tertiary-care hospitals in Tabriz, Iran, in the post-COVID-19 era, characterize the antibiotic resistance phenotypes of these isolates, assess the frequency and distribution of blaTEM and blaCTX-M genes, and statistically evaluate whether the co-occurrence of these genes exceeds what would be expected by chance, which would be consistent with (though not sufficient to prove) plasmid-mediated co-carriage. The findings offer recent post-pandemic molecular epidemiological data from this important region and carry direct implications for antimicrobial stewardship and infection control policies.

Materials and Methods

Bacterial Isolates and Identification

A cross-sectional descriptive study was conducted from December 2024 to November 2025 at three major tertiary-care hospitals in Tabriz, Iran: Imam Reza Hospital, Imam Sajjad Hospital, and Mahallati Hospital. The study was approved by the Research Ethics Committee of the Islamic Azad University, Tabriz Branch (Ethics Code: IR.IAU.TABRIZ.REC.1404.056), and conducted in accordance with the Declaration of Helsinki.

Midstream urine specimens were collected from patients hospitalized in intensive care units (ICUs) with clinical signs of UTI (dysuria, frequency, suprapubic pain, fever). One hundred non-duplicate E. coli isolates recovered from positive urine cultures (≥10⁵ CFU/mL) were enrolled. Identification was performed by standard microbiological methods, including Gram staining, catalase and oxidase tests, growth on MacConkey agar and eosin methylene blue (EMB) agar, and IMVIC biochemical tests. Specimens yielding negative cultures or non-E. coli organisms were excluded.

A “non-duplicate” isolate was defined as the first Escherichia coli isolate recovered from a given patient during the study period; any subsequent isolate from the same patient within the same or a recurrent episode was excluded. The target sample size of 100 isolates was a conventional/practical target for the study period rather than derived from a formal a priori power calculation. Post hoc, a sample of 100 provides adequate precision for the observed 48% ESBL prevalence (eg, a 95% CI half-width of approximately ±10% at this prevalence with n = 100), which is presented here as a retrospective check on adequacy rather than as the original basis for the sample size.

As shown in the patient/specimen flow diagram (Figure 1), 164 urine specimens were submitted from ICU patients during the study period; 29 were culture-negative, leaving 135 culture-positive specimens (all organisms), of which 126 grew Escherichia coli. Of these, 6 were excluded for contamination or polymicrobial growth, and a further 20 were excluded as duplicate isolates (repeat isolates from the same patient), leaving the final 100 non-duplicate E. coli isolates analyzed in this study. Data on prior hospitalization, urinary catheterization, prior antibiotic exposure, and recurrent UTI were not available in the medical records for this cohort and could not be incorporated into isolate classification; this is noted as a limitation (see Discussion, Limitations).

Figure 1.

Patient/specimen flow diagram. Of 164 urine specimens submitted from ICU patients during the study period, 29 were culture-negative, leaving 135 culture-positive specimens (all organisms), of which 126 grew Escherichia coli. Six were excluded for contamination or polymicrobial growth, and a further 20 were excluded as duplicate isolates from the same patient, leaving 100 non-duplicate E. coli isolates for analysis.

Antibiotic Susceptibility Testing

Disk diffusion (Kirby-Bauer method) was performed on Mueller-Hinton agar according to CLSI 2024 guidelines. Antibiotic disks tested included: Cefotaxime (CTX, 30 μg), Ceftazidime (CAZ, 30 μg), Cefepime (FEP, 30 μg), Imipenem (IPM, 10 μg), and Meropenem (MEM, 10 μg). Escherichia coli ATCC 25922 was used as the quality control strain.

Phenotypic ESBL Detection

ESBL production was confirmed phenotypically using the combined disk test (CDT), not a double-disk synergy test, as per CLSI recommendations. Ceftazidime (30 μg) and Ceftazidime/Clavulanic acid (30/10 μg) disks were placed 25 mm apart (center-to-center) on the same Mueller-Hinton agar plate inoculated with a bacterial suspension equivalent to a 0.5 McFarland standard, and their inhibition zone diameters were compared directly on that single plate, rather than a synergy pattern between separated disks. Plates were incubated at 35°C for 16–18 h in ambient air, consistent with CLSI M100 standard incubation conditions for disk diffusion/combined disk testing. Commercially available, CLSI-compliant antimicrobial susceptibility test disks were used. Antimicrobial disks were obtained from Padtan Teb (Padtan Teb Co., Iran). Escherichia coli ATCC 25922 was used as the ESBL-negative quality-control strain and Klebsiella pneumoniae ATCC 700603 as the ESBL-positive quality-control strain for the CDT, per CLSI M100 recommendations. An increase of ≥5 mm in the inhibition zone diameter with the combination disk compared to the single disk was interpreted as ESBL-positive. Cefotaxime-based confirmatory testing was not additionally performed; this is noted as a limitation, since CTX-M-producing isolates may be more readily detected by cefotaxime-based confirmation given the 100% cefotaxime resistance observed in this cohort (see Discussion, Limitations).

DNA Extraction and Genotypic Detection by PCR

Genomic DNA was extracted by the boiling lysis method: 5–6 colonies were suspended in 100 μL of 50 mM NaOH and incubated in a boiling water bath at 100°C for 30 minutes, followed by the addition of 50 μL of 20 mM Tris buffer and centrifugation at 12,000 rpm for 5 minutes. DNA purity and concentration were assessed by NanoDrop spectrophotometry (A260/A280 ratio: 1.8–1.9).

Specific primers were employed for amplification of blaTEM (forward: 5’-GGTGAAAGTAAAAGATGCTGAAG-3’; reverse: 5’-AACTTTATCCGCCTCCATC-3’; amplicon 559 bp) and blaCTX-M (forward: 5’-GCGGAAAAGCACGTCAATGGG-3’; reverse: 5’-GCCAGATCACCGCGATATC-3’; amplicon 427 bp). PCR reactions (20 μL total volume) contained 10 μL 2× Green Master Mix (Taq DNA polymerase), 0.5 μL of each primer, 1 μL template DNA, and 8 μL sterile deionized water. Cycling conditions for blaTEM: 95°C/3 min initial denaturation; 35 cycles of 95°C/30 s, 55°C/30 s, 72°C/60 s; 72°C/5 min final extension. For blaCTX-M: annealing at 60°C. Amplicons were resolved by 1.5% agarose gel electrophoresis, stained with DNA Safe Stain, and visualized under UV transillumination.

The positive control for both blaTEM and blaCTX-M PCR reactions was a clinical E. coli isolate previously confirmed to carry the respective gene by Sanger sequencing.

Statistical Analysis

Data were analyzed using IBM SPSS Statistics for Windows, Version 26.0 (IBM Corp., Armonk, NY, USA). Frequencies and percentages were calculated for categorical variables. The statistical independence of the co-occurrence of blaTEM and blaCTX-M was evaluated in a 2×2 contingency table (blaTEM positive/negative × blaCTX-M positive/negative). Fisher’s exact test was used as the primary statistic given expected cell counts below 5 in the table, with Pearson’s chi-square test (with and without Yates’ continuity correction) reported as a secondary/sensitivity measure. The odds ratio (OR) with 95% confidence interval, calculated by the Woolf/log method, was used as the primary effect-size estimate. Expected co-occurrence under independence was calculated as: P(blaTEM) × P(blaCTX-M) × n. 95% confidence intervals (CIs) for proportions were computed using the Wilson method. Gene frequencies were additionally compared across the three participating hospitals to assess for hospital-level clustering. P-values <0.05 were considered statistically significant.

Results

Patient Demographics and Species Distribution

Of 135 culture-positive urine specimens from ICU patients, 100 (74.07%) yielded E. coli as the causative pathogen, followed by Klebsiella pneumoniae (11.85%), Proteus mirabilis (5.92%), Enterococcus faecalis (4.44%), and Pseudomonas aeruginosa (3.73%). The mean patient age was 47 ± 16 years, with females comprising 90% of the UTI cohort.

Antibiotic Susceptibility Profiles

Antibiotic susceptibility testing of the 100 E. coli isolates revealed complete resistance to Cefotaxime (100%) and near-universal resistance to Ceftazidime (95%). Intermediate resistance patterns were observed for Cefepime (44% resistant, 6% intermediate, 50% susceptible). Carbapenem activity was largely preserved, with 70% susceptibility to Meropenem and 82% susceptibility to Imipenem. Full susceptibility data are presented in Table 1.

Table 1.

Antibiotic Susceptibility of E. coli Isolates (n = 100)

Antibiotic	Susceptible (%)	Intermediate (%)	Resistant (%)
Cefotaxime (CTX)	0	0	100
Ceftazidime (CAZ)	5	1	94
Cefepime (FEP)	50	6	44
Meropenem (MEM)	70	12	18
Imipenem (IPM)	82	8	10

Phenotypic ESBL Production

Combined disk testing confirmed ESBL production in 48 of 100 (48.0%; 95% CI: 38.2–57.9%) E. coli isolates. All ESBL-producing strains demonstrated resistance to at least one third-generation cephalosporin, consistent with the phenotypic definition.

Genotypic Characterization of ESBL Genes

Among the 48 phenotypically confirmed ESBL-producing isolates, PCR analysis revealed: blaTEM in 28 (58.3%; 95% CI: 44.4–72.3%), blaCTX-M in 20 (41.7%; 95% CI: 27.7–55.6%), and co-carriage of both genes in 18 isolates (37.5%; 95% CI: 23.8–51.2%). Broken down into four mutually exclusive categories, 10 isolates (20.8%) carried blaTEM only, 2 isolates (4.2%) carried blaCTX-M only, 18 isolates (37.5%) carried both genes, and 18 isolates (37.5%) carried neither gene (10 + 2 + 18 + 18 = 48). These genotypic frequencies are summarized in Table 2. The amplification of Gene blaTEM and Gene blaCTX-M was confirmed by PCR, as shown in Figures 2 and 3, respectively.

Table 2.

Genotypic Distribution of ESBL Genes Among Confirmed ESBL-Producing E. coli (n = 48)

Gene	n	Frequency (%)	95% CI (%)
blaTEM	28	58.3	44.4–72.3
blaCTX-M	20	41.7	27.7–55.6
Both blaTEM + blaCTX-M	18	37.5	23.8–51.2
blaTEM only (no blaCTX-M)	10	20.8	11.7–34.3
blaCTX-M only (no blaTEM)	2	4.2	1.2–14.0
Neither gene	18	37.5	25.2–51.6

Figure 2.

PCR amplification of the blaTEM gene. Lane L100 bp DNA ladder; Lane PC: positive control (a sequence-confirmed clinical isolate previously confirmed positive for blaTEM by Sanger sequencing, 559 bp); Lanes 3–10: clinical isolates positive for the blaTEM gene (559 bp); Lane NC: negative control.

Figure 3.

PCR amplification of the blaCTX-M gene. Lane L 100 bp DNA ladder; Lane PC: positive control (a sequence-confirmed clinical isolate previously confirmed positive for blaCTX-M by Sanger sequencing, 427 bp); Lanes 3, 4, and 6–10: clinical isolates positive for the blaCTX-M gene (427 bp); Lane NC: negative control.

Statistical Analysis of blaTEM and blaCTX-M Co-Occurrence

The expected frequency of co-carriage under the null hypothesis of statistical independence was calculated as: 0.583×0.417 × 48 = 11.67 isolates. The observed co-carriage of 18 isolates markedly exceeded this expectation. Using the full 2×2 contingency table (both = 18, blaTEM only = 10, blaCTX-M only = 2, neither = 18), Fisher’s exact test—the primary statistic given expected cell counts below 5—yielded p = 0.000255. Pearson’s chi-square test with Yates’ continuity correction yielded χ² = 12.0 (df = 1, p = 0.000532), and the uncorrected chi-square was 14.15 (p = 0.000169), consistent with the exact test. The odds ratio for co-occurrence was 16.2 (95% CI: 3.10–84.59), indicating a strong, statistically significant, non-random association between the two genes. These statistical findings are consistent with, but do not establish, co-localization of blaTEM and blaCTX-M on a common plasmid backbone; PCR detection alone cannot determine whether the two genes reside on the same plasmid, on different plasmids, or on the chromosome, nor can it distinguish horizontal transfer from clonal expansion of a successful lineage. Confirmatory methods such as plasmid replicon typing, conjugation/transformation assays, S1-PFGE with Southern hybridization, or whole-genome sequencing would be required to establish physical linkage or co-transfer, and were not performed in this study (see Limitations). The complete statistical summary is presented in Table 3.

Table 3.

Statistical Analysis of blaTEM and blaCTX-M Co-Occurrence Among ESBL-Producing E. coli (n = 48)

Parameter	Value
blaTEM only (no blaCTX-M), n	10
blaCTX-M only (no blaTEM), n	2
Neither gene, n	18
Observed co-carriage (n)	18
Expected co-carriage under independence	11.7
Chi-square, Yates-corrected (df=1)	12.0
Chi-square p-value, Yates-corrected	0.000532
Fisher’s exact p-value	0.000255
Chi-square, uncorrected (df=1)	14.15
Chi-square p-value, uncorrected	0.000169
Odds ratio (95% CI)	16.2 (3.10–84.59)
Interpretation	Significant non-random co-occurrence (p < 0.001); Fisher’s exact test used as primary statistic given small expected cell counts

Discussion

This study provides molecular epidemiological data on ESBL-producing E. coli in ICU-associated UTIs from Tabriz, northwestern Iran, during the post-COVID-19 era. Four principal findings with significant epidemiological and clinical implications emerged from our investigation.

The ESBL prevalence of 48% among E. coli UTI isolates in this study is substantially higher than rates reported in earlier Iranian investigations. Miraalami et al (2015) reported ESBL prevalence of approximately 30–40% in Tabriz before 2015,¹³ while a national study by Halaji et al (2020) reported 45% ESBL prevalence among kidney transplant patients in Isfahan.¹⁴ Our elevated prevalence aligns with post-pandemic reports from neighboring countries, including Turkey (52%)¹⁵ and Iraq (56%),¹⁶ consistent with the hypothesis that antibiotic overuse during the COVID-19 pandemic contributed to ESBL dissemination regionally; however, this study lacks a pre-pandemic comparison cohort from the same hospitals, so a direct causal link cannot be established from these data alone. The universal cefotaxime resistance (100%) and near-universal ceftazidime resistance (95%) observed in our cohort indicate very limited utility for these agents as empiric therapy in this setting, based on the observed resistance rates.

The dominance of blaTEM (58.3%) over blaCTX-M (41.7%) in our isolates presents a noteworthy contrast to contemporary global reports, wherein blaCTX-M has been recognized as the predominant ESBL type since the early 2000s.⁶ Several explanations are plausible: local antibiotic prescription patterns favoring aminopenicillins and first-generation cephalosporins—the classical selective pressures for blaTEM—may persist in this clinical environment; alternatively, this finding may reflect differential representation of clonal lineages circulating in Tabriz ICUs. Studies from Iraq¹⁶ and Sudan¹⁷ similarly reported blaTEM as the dominant gene type in UTI isolates, contrasting with CTX-M dominance in Western Europe and North America.⁶ This geographic heterogeneity in ESBL gene prevalence reinforces the need for region-specific genotypic surveillance rather than extrapolation from global datasets.

The most epidemiologically significant finding of this study is the statistically confirmed non-random co-occurrence of blaTEM and blaCTX-M. The observed co-carriage of 18 isolates (37.5%) compared to the expected 11.7 under independence (p < 0.001 by both Chi-square and Fisher’s exact tests) demonstrates a strong statistical association between the two genes, which is consistent with, but does not by itself establish, plasmid-mediated co-carriage as a possible dissemination mechanism—PCR alone cannot confirm physical linkage or distinguish it from clonal spread (see Limitations). This phenomenon has been increasingly documented in ESBL-producing Enterobacterales globally, where IncF-type conjugative plasmids often simultaneously harbor blaTEM, blaCTX-M, and additional resistance determinants, including quinolone resistance genes.^8,18 The clinical and infection control implications warrant attention: surveillance programs relying on a single gene target may underestimate the reservoir of dual ESBL carriers, and interventions targeting one gene alone may not address isolates that also carry the partner gene. The data presented here therefore argue for simultaneous, dual-target genotypic surveillance in high-risk hospital units.

From a therapeutic standpoint, the preserved carbapenem susceptibility—Imipenem 82%, Meropenem 70%—suggests carbapenems retained the highest in vitro activity among the agents tested in this cohort; this cannot be generalized into a full empirical-therapy recommendation without a broader susceptibility panel (see Limitations) and patient-level clinical outcome data. The relatively low but non-trivial rates of carbapenem resistance (10–18%) are concerning and may reflect co-production of carbapenemases in a subset of isolates, although confirmatory testing (MIC determination and carbapenemase-gene screening) was beyond the scope of this study and is recommended before drawing firm conclusions. The absence of pan-resistance is consistent with, but does not confirm, a lower burden of carbapenem-resistant Enterobacterales (CRE) in this population compared to some neighboring countries;⁴ confirmatory testing would be needed to establish this. Nonetheless, rigorous carbapenem stewardship is warranted to help prevent progression toward CRE dominance.

Compared with the largest published Iranian series, Nejat et al (2024) reported 69.6% ESBL prevalence and 88.5% blaCTX-M frequency in Ahvaz,¹⁹ while Sarshar et al (2021) documented 44% ESBL prevalence in Klebsiella pneumoniae UTI isolates in Hashtgerd, with 34% blaCTX-M and 28% blaTEM.²⁰ The lower blaCTX-M frequency in our cohort compared to Ahvaz may reflect distinct regional epidemiology or earlier-phase CTX-M displacement in Tabriz. International comparisons reveal congruence with Egyptian data (Hassuna et al, 2020: 59.7% ESBL, blaTEM 75%), Indian data (Chaudhary et al, 2023: blaTEM 63.4%, blaCTX-M 36.6%), and Sudanese data (Dirar et al, 2020: blaTEM 86%, blaCTX-M 78%), all of which document blaTEM as the more prevalent gene, in contrast to European reports.^17,21,22

Several limitations of this study merit acknowledgment. The sample was restricted to ICU settings at three hospitals in a single city, potentially limiting generalizability to community-acquired infections and other geographic regions of Iran. Data on prior hospitalization, urinary catheterization, prior antibiotic exposure, and recurrent UTI were not available in the medical records for this cohort, precluding a more detailed characterization of healthcare-associated risk factors; the target sample size was also set as a practical/conventional target for the study period rather than derived from a formal a priori power calculation (see Methods, Bacterial Isolates and Identification). The boiling lysis DNA extraction method, while cost-effective and widely used, may yield lower-quality DNA compared to commercial kit-based methods, potentially affecting PCR sensitivity. Subtyping of CTX-M variants (eg, CTX-M-1, CTX-M-15, CTX-M-27) was not performed, precluding phylogenetic inference on the circulating CTX-M clade. Furthermore, plasmid typing and conjugation experiments were not conducted to directly confirm co-transfer and PCR detection alone cannot distinguish horizontal gene transfer from clonal expansion of a successful lineage (eg, ST131); the statistical co-occurrence reported here is consistent with, but does not establish, either mechanism. In addition, the antibiotic panel tested was restricted to five β-lactams relevant to ESBL phenotype confirmation; commonly used non-β-lactam agents (fluoroquinolones, aminoglycosides, trimethoprim-sulfamethoxazole, nitrofurantoin, fosfomycin) and co-resistance markers were not assessed, so this study cannot fully characterize multidrug resistance or support a complete empirical-therapy recommendation. The universal cefotaxime resistance observed alongside a lower phenotypic ESBL rate (48%) may partly reflect AmpC β-lactamase production, porin loss or efflux-related resistance, or limitations of combined-disk phenotypic testing in distinguishing ESBL from other cephalosporinase mechanisms; AmpC-specific testing (eg, cefoxitin screening) was not performed and is recommended in future work. Future studies incorporating whole-genome sequencing (WGS) and long-read sequencing platforms would enable comprehensive characterization of plasmid backbones, ESBL gene environments, and clonal lineages in this region.

Conclusions

This study demonstrates that ESBL-producing E. coli are highly prevalent (48%) in ICU-associated UTIs in Tabriz, Iran, with blaTEM (58.3%) emerging as the dominant ESBL gene type in this post-pandemic cohort. The statistically significant co-occurrence of blaTEM and blaCTX-M genes (observed: 37.5%; expected under independence: 24.4%; p < 0.001) is consistent with, though does not by itself establish, plasmid-mediated co-carriage as a possible dissemination mechanism, with implications for surveillance design that would benefit from simultaneous dual-gene genotyping. Universal cefotaxime resistance renders this drug of limited or no utility for empirical therapy in this setting, based on the observed resistance rate, while preserved carbapenem activity in this cohort suggests these agents remain among the more reliable options for empirical therapy locally; this cannot be generalized into a formal treatment recommendation without a broader antibiotic panel and patient-level outcome data (see Limitations). Urgent implementation of enhanced antimicrobial stewardship programs, contact precautions for ESBL-positive ICU patients, and routine molecular surveillance that incorporates both blaTEM and blaCTX-M targets are recommended to help monitor for progression toward a CRE-dominated epidemiology in northwestern Iran.

Funding Statement

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Ethical Considerations

This research was carried out in full accordance with the ethical guidelines established by the institutional research committee, as well as the Declaration of Helsinki (1964) and its subsequent revisions, or equivalent ethical standards. Approval for the study protocol was granted by the Ethical Committee of Islamic Azad University, Tabriz Branch (Ethical code: IR.IAU.TABRIZ.REC.1404.056). Before specimen collection, informed consent was secured from every participant enrolled in the study. Each participant received information about the research aims and procedures, along with an explanation of their right to withdraw from the study at any time.

Author Contributions

MS: Conceptualization, Data curation, Formal analysis, Investigation, Methodology. SHA: Formal analysis, Investigation, Writing—review and editing.

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no conflict of interest.