Exploring the genomic landscape of antimicrobial resistance in XDR-ESKAPE pathogens: evidence-based implications for organism-targeted therapy in sepsis

1 Introduction

Sepsis is a syndrome of multiple organ dysfunction triggered by a dysregulated host response to an infection and is usually fatal. It is a key contributor to global mortality in patients with sepsis (1). Most sepsis patients receive empiric antibiotic therapy upon admission before culture results are available. However, this approach may be ineffective in upto 40% of the cases due to the diversity of causative pathogens and their varied inherent or acquired antimicrobial resistance (AMR) mechanisms (2). In sepsis, a major factor contributing to poor prognosis is the emergence of extremely drug-resistant (XDR) organisms, particularly those classified within the ESKAPE group (Enterococcus faecalis, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Escherichia coli). The ESKAPE pathogens present a significant challenge in the case of sepsis, as their rapid progression from colonisation to invasive infection, combined with extensive resistance profiles, frequently results in initial antibacterial treatment failure and poorer clinical outcomes as they elude the inhibiting effect of different antibiotic classes.

Thus, identifying and defining the plausible mechanism(s) adapted by the ESKAPE pathogens early on in the treatment of sepsis is essential for guiding antibiotic therapy in the absence of a confirmed microbiological diagnosis. Unavoidably, this empirical use of broad-spectrum antibiotics in the initial stage contributes to an increased AMR burden in sepsis patients due to multiple resistance mechanisms adopted by causal bacteria, against antibiotics commonly used in ICUs (3). Antibiotic Stewardship Programs (ASPs) generally rely on institutional antibiograms that provide generalised susceptibility data and may not reflect the real-time resistance patterns of specific pathogens infecting individual patients. This gap is especially serious in cases involving XDR-ESKAPE, where genetic determinants often go undetected in conventional antibiotic susceptibility testing (AST) workflows (4).

Although whole-genome sequencing (WGS) has significantly advanced the characterisation of antimicrobial resistance in bacterial pathogens (5), no study has yet provided a comprehensive functional AMR gene profile for the entire XDR-ESKAPE group, specifically in reference to sepsis. Previous WGS-based investigations have primarily focused on individual species, selected resistance determinants, non-sepsis clinical sources, resulting in a critical gap in defining the resistome landscape of XDR-ESKAPE pathogens in sepsis and effective clinical management of these XDR infections.

The present study addresses this gap by characterising the resistance profiles of XDR-ESKAPE pathogens isolated from patients with sepsis using whole-genome sequencing (WGS) and proposes a novel approach to sepsis management using organism targeted combination therapy. This approach enables correlation of genotypic and phenotypic resistance profiles, and proposes in vitro evidence-based synergistic antibiotic-combination selection curated through genomic mapping.

2 Methods

2.1 Sample collection, isolation and identification of morphotypes

The study enrolled a total of 100 patients with confirmed sepsis diagnosis (SOFA score ≥2), admitted in the general ICU, CCU, and pulmonary wards of Bhopal Memorial Hospital and Research Centre, Bhopal. Written informed consent was obtained from all subjects before enrolment in accordance with the approval of the Institutional ethical committee (IEC/46/Research/20).

3 mL of whole blood was collected aseptically in an EDTA vacutainer, and/or 0.5 mL of Broncho-alveolar lavage (BAL) fluid was collected in a sterile BAL fluid container. Blood culture is commonly performed using the BACTEC blood culture system which requires a minimum of 10 mL of blood sample for diagnosis, a challenging process for patients suffering from sepsis. Therefore, for the present study, the following procedure was followed. Standard bacterial species (Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterococcus faecalis) were procured from ATCC for reference. A culture was inoculated for each of the species in trypticase soy broth (TSB) and kept for incubation for 24 h at 37 °C. To standardize the technique of using a minimum required amount of blood for culture, O + blood was obtained from a voluntary donor and the blood was then inoculated with 10 μL of the prepared inoculum of standard bacterial species. The inoculated blood was divided into three culture flasks with the following ratios: 1 mL: 10 mL, 3 mL: 30 mL, and 5 mL: 50 mL blood to TSB. All three flasks were kept in a shaking incubator for 7 days at 37 °C. All three samples were spread on to blood agar plates on the second and seventh day of incubation to observe bacterial growth (6, 7). The results were compared with the BACTEC results of the same sampling (concordance rate 94.45%). The BAL fluid was cultured by adding 1 mL of TSB to 0.5 mL of BAL fluid and incubating overnight at 37 °C in a shaking incubator (8). The cultured inoculum was spread onto trypticase soy agar (TSA) plates after incubation and incubated for an additional 24 h at 37 °C to assess bacterial growth. Each organism was isolated in its pure form, and standard biochemical tests, including Gram staining, were performed to identify the bacterial genus (9).

2.2 Species identification of morphotypes by VITEK 2.0

The biochemical profiles of the isolates were confirmed using the VITEK 2 system (BioMérieux, France) according to the manufacturer's instructions to identify the species of the morphotypes. Pure colony of each morphotype was cultured on TSA, then suspended in sterile saline (0.9% NaCl) to a turbidity of McFarland 0.5–0.63. Then, 0.5 mL of the bacterial inoculum was injected into Gram-positive (GP) and Gram-negative (GN) cards to run 64 tests combinations (10). Each test result was then matched in the VITEK database; wherein bacterial species were identified with a 99% similarity.

2.3 Antibiotic susceptibility assessment for antibiotic tolerance

The disc diffusion method was performed to screen the identified isolates for primary antibiotic tolerance (11). The antibiotic classes chosen for the study were divided into two panels: the first panel comprised of the general antibiotic regimen prescribed for bacterial infections (12). The second panel consisted of antibiotics prescribed in the institute's hospital (13) for the treatment of sepsis (Supplementary Table S1). The antibiotic discs (HiMedia, India) of each antibiotic were placed on Mueller-Hinton Agar (MHA) plates containing a bacterial lawn of individual isolates, prepared by spreading 1 mL of bacterial inoculum (OD of 0.6) on a 90 mm MHA plate (14). The plates were incubated for 24 h at 37 °C to obtain a zone of inhibition, which was noted and measured as per CLSI guidelines (15) and EUCAST guidelines (16).

2.4 Establishing antibiotic resistance patterns in tolerant isolates

To determine the resistance pattern of the isolates against the panel of antibiotics, a modified version of the screening method described by Liu et al. (17) was used. The isolates were grown on MHA containing each antibiotic at the disc concentration to which they were tolerant. They were then sub-cultured for five generations and subsequently regrown on MHA without antibiotics for three generations. Antibiotic stress was again induced by growing the isolates on MHA containing the relevant antibiotic concentration for 24 h at 37 °C. The results were recorded as visible colony growth after each generation (subculture). Isolates found to be resistant to more than 5 classes of antibiotics were grouped as XDRs (18).

2.5 Minimum inhibitory concentration of XDRs

The minimum inhibitory concentration refers to the least concentration of the antibiotic that effectively inhibit visible growth of the bacterial isolate. The XDR-ESKAPE isolates were subjected to the micro-broth dilution method (19), where 100 μL of pre-prepared inoculum (1.5 × 10⁶ CFU/mL) in TSB was added to micro-centrifuge tubes containing 500 μL of TSB resuspended with antibiotics ranging from 0.01 mg/mL to 40 mg/mL. The tubes were incubated at 37 °C in a shaking incubator for 18 h. A negative control (TSB) and a tube with equal volume of TSB with inoculum were also incubated. The MIC value was established by recording absorbance of the incubated samples at 595 nm in a microplate reader (iMark, Biorad).

2.6 Whole genome sequencing of XDRs

2.6.1 WGS sample size and selection criteria

Minimum Inhibitory concentrations (MIC) for antibiotics selected from each of the 7 antibiotic classes (refer section 1.1.3) were calculated and was set as the selection criterion for XDR-isolates table (Supplementary Table S2). The strains/ isolates within each species which demonstrated the highest MIC values against ≥4 antibiotic classes were chosen for whole genome sequencing; therefore 12 XDR-ESKAPE isolates were taken forward for analysis and in vitro testing for combination therapy.

2.6.2 Illumina HiSeq based whole genome sequencing

Whole-genome sequencing of the XDR-ESKAPE isolates (isolates against which antibiotics had the highest MIC value; n = 12) was performed to identify the genes involved in their antibiotic resistance mechanism. The DNA of XDR-ESKAPE was extracted using the PurelinkGenomic DNA kit (Invitrogen, India), quantified using QUBIT 3 fluorometer (ThermoFisher Scientific, US), and purified using AMPure XP beads (Beckman Coulter, US). The DNA was then fragmented into ∼350 bp using an ultrasonicator. The fragmented DNA was end-repaired, ligated, and then subjected to PCR for amplification using Illumina universal primers. Sequencing was performed on the Illumina HiSeq 2,500 platform generating 1GB of data per sample. The Unicycler software (GPLv3.0.) (20) was primarily used to assemble the generated raw data. It was reassembled using the corresponding sequences with the help of MeDuSa (version 1.6) (21). The reference genome accession no. used in this study were Escherichia coli – LT601384.1; Staphylococcus aureus – NC_007795.1; Klebsiella pneumoniae – HF536482.1; Acinetobacter baumannii – NZ_CP043953.1; Pseudomonas aeruginosa – NC_002516.2; Enterococcus faecalis – CP124778.1. The assembled sequences were then annotated using PROKKA (version 14.6). Molecular Evolutionary Genetics Analysis (MEGA v12.1) was used to analyze the genomes by maximum-likelihood (ML) algorithm and construction of the phylogenetic trees. The circular maps of all XDR-ESKAPE organisms were created using CGView software (version 1.0) (22). The assembled FASTA DNA files were analysed using the antibiotic resistance database (AMRFinder plus, version 3.2) to identify the functional antibiotic resistance genes present in the XDR-ESKAPE genomes and the possible mechanism(s) of resistance.

2.7 Fractional inhibitory concentrations of the antibiotic combinations

The fractional inhibitory concentration (FIC) was assessed to check whether the antibiotic combinations tested in the study after the AMR gene analysis can be used synergistically. The MIC, MIC/2 and MIC/5 concentrations of ineffective antibiotics were used in combination with the concentrations of the potentiating antibiotics.100 μL of 1.5 × 10⁶ CFU/mL of XDR ESKAPE culture was added to each of the test combinations and the cultures were incubated at 37 °C in a shaking incubator for 18 h. A negative control (TSB) and a tube with equal volume of TSB with inoculum were also incubated. The absorbance was measured at 595 nm and FIC was calculated as the ratio of the MIC of the antibiotic in combination to the MIC of the individual antibiotic (19). The experiment was conducted in duplicates and replicated twice.The interactions were classified as synergistic (if ΣFIC ≤ 0.5), additive (if ΣFIC > 0.5-1), indifferent (if ΣFIC > 1-4) and antagonistic (if ΣFIC > 4.0) (23).

FICA=MICofAntibiotic1incombinationMICofAntibiotic1alone

FICB=MICofAntibiotic2incombinationMICofAntibiotic2alone

ΣFIC=FICA+FICB

A complete workflow of the methodology followed for this study is diagrammatically explained in Figure 1.

2.8 Statistical analysis

All the data collected in the study were plotted using GraphPad Prism version 8.0.1 and analysed statistically. A p-value < 0.05 was considered statistically significant.

3 Results

3.1 Distribution pattern of isolates

In this study, a total of 43 distinct bacterial morphotypes were isolated from the clinical samples of bacterial sepsis patients (blood and BAL fluid). Out of these, 36 (83.72%) were identified as ESKAPE organisms using VITEK 2.0. In the ESKAPE group, Escherichia coli was the most predominant species, followed by Klebsiella pneumoniae, Enterococcus faecalis, Pseudomonas aeruginosa, and Staphylococcus aureus (Table 1).

Table 1. Distribution and frequency of the bacterial species (ESKAPE organisms and other species) isolated from sepsis samples.

Notably, only 4 strains (9.3%) of Acinetobacter baumannii were isolated. Bacterial species other than ESKAPE, including Staphylococcus haemolyticus, Stenotrophomonas maltophilia, Moraxella lacunata, and several Enterobacter species (E. cloacae, E. hormaechei, E. aerogenes) were also isolated (16.3%) (Table 1).

3.2 Antibiotic resistance pattern

ESKAPE organisms are the leading cause of antibiotic-resistant bacterial sepsis, significantly limiting treatment options and increasing mortality. Since this investigation observed a high prevalence of ESKAPE organisms, their antibiotic resistance profiling was performed. The profile revealed antibiotic resistance across multiple drug classes with A. baumannii and P. aeruginosa being the predominant contributors to the overall resistance pool because all their strains (refer Table 1) were found resistant to maximum number of antibiotic classes, including β-lactams, fluoroquinolones, glycopeptides, macrolide, nitrofuran, oxazolidinone, and lincomycin (AZM, A/S, CXM, NIT, AMC, CD, CIP, LZ, and VA). K. pneumoniae also showed substantial resistance, with 3 of the 4 XDR strains resistant to 13 antibiotics, while the other single strain was MDR, only resistant to AZM, CXM, and fluoroquinolones. S. aureus strains (2) displayed susceptibility towards IPM, OF and LE, but were found resistant to macrolides, β-lactams, nitrofuran, lincomycin, oxazolidinone, and glycopeptides. Among E. coli strains, 7 showed a moderate-to-high resistance pattern, with all 7 strains resistant to all seven antibiotic classes. All of the E. faecalis strains showed resistance against β-lactams, fluoroquinolones, glycopeptides, macrolide, nitrofuran, oxazolidinone, and lincomycin. Overall, the findings highlight high resistance in ESKAPE organisms, particularly in A. baumannii, S. aureus, and E. coli strains, while E. faecalis demonstrated relatively lower resistance. However, all isolates were found to be susceptible to meropenem (Figure 2).

3.3 Minimum inhibitory concentration of XDR-ESKAPE

The MIC values of all screened antibiotic classes in the panel were assessed using micro-broth dilution method, where each strain within a species (ESKAPE group) showed a different trend in MIC values. The strains that recorded the highest MICs in all antibiotic classes were SB13, SB14, SB68 (E. coli strains), SB54, SB57 (S. aureus strains), SB26, SB28 (K. pneumoniae strains), SB44, SB47 (A. baumannii strains), SB36, SB40, SB72 (P. aeruginosa strains), SB09 and SB43 (E. faecalis strains). The MIC values have been presented in Supplementary Table S2.

3.4 Antibiotic-resistant genes in ESKAPE

A comprehensive summary of the DNA blueprint of the ESKAPE group with their intrinsic and functional AMR genes was obtained from the WGS sequenced data using the antibiotic resistance database (AMRFinder plus, version 3.2). Following the phenotypic resistance profiling, WGS of the 12 identified XDR-ESKAPE strains was conducted to gain an insight into the genetic basis of the antibiotic resistance pattern(s) and underlying mechanisms that drive the phenotypic resistance. The genes mtrA, NmcR, marA, rpoB, and optrA were commonly found in all XDR-ESKAPE strains, although their frequencies varied among the strains. The analysis also identified a unique pattern of AMR genes specific to a particular bacterial species conferring intrinsic resistance within the ESKAPE group (Table 2).

Table 2. Table depicting the specific genes which confer intrinsic resistance against specific antibiotics in the studied ESKAPE organisms.

The presence of specific intrinsic resistance AMR genes among ESKAPE organisms isolated from blood and BAL-F (Table 2) strongly discourages the use of the certain antibiotic classes, i.e., β-lactams and fluoroquinolones, as empirical therapy for sepsis. β-lactams, including ampicillin/sulbactam and imipenem, may prove unreliable due to the presence of marA gene in E. coli, SHV-1 and LptD genes in K. pneumoniae, adeL, adeK, AbaQ genes in A. baumannii, and mexP, mexK, and mexF genes in P. aeruginosa. Similarly, fluoroquinolones (ciprofloxacin, levofloxacin, and ofloxacin) should not be preferred for initial therapy due to the presence of active efflux pump genes such as acrF in E. coli, norA in S. aureus, and efrA in E. faecalis. Therefore, empirical use of the antibiotics for sepsis treatment should be done very diligently as it would carry a high risk of therapeutic failure, delayed effective treatment, and increased mortality, as evident from the results presented above.

Genotyping of AMR strains also revealed the presence of other functional resistance genes in these 12 strains of XDR-ESKAPE (Table 3), wherein the extensive phenotypic resistance corroborates with the corresponding AMR genes. The most common genes among the gram-negative organisms of XDR-ESKAPE were found to be bla_OXA, bla_TEM−1, and bla_CTX−M−15 genes, conferring resistance to β-lactams and the cephalosporin class of antibiotics, along with tet(A) and tet(B) against tetracyclines. Active resistance genes against fluoroquinolones (qnrB, gyrA, parC/parE), nitrofuran (OqxA) and lincomycin (ermC) were also present. Each of these strains showed the presence of multiple active efflux pumps (such as emrD, KdeA, acrF, adeC, amvA, mex_A/E/X). A compilation of all functional genes present in the XDR-ESKAPE is presented in Figure 3.

Table 3. Phenotypic resistance profiles and the corresponding major antimicrobial resistance (AMR) genes in sequenced XDR-ESKAPE isolates.

AZM, azithromycin; A/S, ampicillin/sulbactam; CXM, cefuroxime; IPM, imipenem; NIT, nitrofurantoin; OF, ofloxacin; AMC, amoxicillin/clavulanate; CD, clindamycin; CIP, ciprofloxacin; LE, levofloxacin; LZ, linezolid; PTZ, piperacillin/tazobactam; VA, vancomycin.

A prominent observation was the presence of functional genes responsible for resistance against antibiotics other than the 14 chosen in the study, which included Fosfomycin (fosA), Streptomycin [ant(3’’)-Ia, aadA2], Sulfonamide (sul1, sul2), gentamycin [aac(6’)-Ib, aac(3)-IIe], and chloramphenicols (catA, catB), indicating that these clinical strains of ESKAPE develop a broad resistome, characterized by the co-existence of multiple functional AMR genes and active efflux pumps. A comparison of the functional AMR genes of each genotyped ESKAPE organism with the respective standard reference genome (Figure 3) highlights the intra-species heterogeneity in the resistome and reveals the acquisition of diverse resistance determinants by XDR-ESKAPE. It also mechanistically explains the observed phenotypic resistance and offers baseline data for local surveillance, molecular diagnostics and treatment designing options.

3.5 Antibiotic combination therapy against sepsis isolates

The findings from the phenotypic and genotypic analysis of clinical isolates of sepsis XDR-ESKAPE in our work provided comprehensive information on AMR genes and their functional resistance mechanisms. In accordance with the demonstrated results above, the presence of multiple active resistance determinants is indicative that they may prevent any of the antibiotics from effectively combating the infection in sepsis caused by these XDR strains. Since certain AMR genes (linked to antibiotics such as meropenem, cefiderocol, polymyxin B, and colistin) were not detected in the studied group, the genotypic analysis suggests that they can be used in monotherapy. However, all of them belong to the seven resistant classes of antibiotics investigated, there is a possibility that monotherapy regimen of even these non-resistant antibiotics may start to fail, which suggests points to the selection of an antibiotic combination as an alternative approach therapy. The FIC index based proposed combinations against the sequenced XDR-ESKAPE isolates are presented in Tables 4, 5, which were selected on the basis of AMR data and the combination mechanism documented in the literature (refer checkerboard FIC raw data in Supplementary Table S3). The evidence helped to design tailored, pathogen-specific combination regimens that can target the resistance mechanism by selecting antibiotics whose resistance genes were absent in the target organism. This strategy can aid in suppression of resistance, enhancement of bacterial eradication, and improvement in patient outcomes in high-risk bacterial sepsis infections caused by XDR-ESKAPE.

Table 4. Proposed combinations of potential synergistic antibiotics evaluated against the resistant gram-negative XDR strains tested in vitro, based on the underlying resistance mechanisms and FIC indices.

FIC, fractional inhibitory concentration; S, synergism; Ad, additivity; Abx, antibiotic.

Gene=as per WGS data presented.

Table 5. Proposed combinations of potential additive antibiotics evaluated against the resistant gram-positive XDR strains tested in vitro, based on the underlying resistance mechanisms and FIC indices.

FIC, fractional inhibitory concentration; Ad, additivity; Abx, antibiotic.

Gene=as per WGS data presented.

4 Discussion

Sepsis is one of the major contributors to global mortality in ICU settings, where rapid identification and prompt treatment are essential in improving the outcomes. The foremost concern that healthcare facilities deal with in the present times is antibiotic-resistant XDR pathogens infecting patients in the ICU, leading to the development of sepsis. The current investigation on sepsis was undertaken to characterize the phenotypic/genomic profile of sepsis-causing extensively resistant bacteria (XDR-ESKAPE), in an attempt to draft an evidence-based framework for development of targeted therapies against specific causal AMR phenotypes, for improving antibiotic stewardship.

In this study, the prevalence pattern of identified XDR-ESKAPE in our tertiary care setting was analysed, in which the XDR E. coli was highest amongst bacterial sepsis samples (Table 1), followed by K. pneumoniae, P. aeruginosa, E. faecalis, and S. aureus. The least isolated species was A. baumannii. A survey by Bashir et al. (24), conducted on isolates from patients with hospital-acquired infections in a tertiary care hospital revealed a similar trend where E. coli was the most prevalent species (29.35%), which was concordant to the result of our work.

Examining the resistance profiles of the isolated XDRs, it was observed that 2 of the 4 isolated A. baumannii strain were resistant to all seven classes of antibiotics, followed by P. aeruginosa (all 5 strains). However, among the E. coli strains, 7 of the total isolated strains were found to be XDR. According to the review of literature by Mancuso et al. (25) A. baumannii was the most reported resistant species in hospital acquired infections, followed by P. aeruginosa, E. coli and S. aureus amongst the ESKAPE group, a finding similar to ours.

The genomes of sepsis-causing XDR-ESKAPE were examined to gain an insight into the functional AMR genes, which revealed the presence of signature genes in each ESKAPE organism conferring intrinsic resistance to fluoroquinolones, β-lactams, and carbapenem antibiotics. The OqxA gene, which is reported to be a part of the OqxAB efflux pump that confers resistance to multiple antibiotics and is specific to E. coli (26), was found in the genomes of the XDR E. coli strains. Given that the OqxA gene is associated with mechanisms that can contribute to resistance against oxazolidinone-class antibiotics, caution is advised, and the use of drugs such as linezolid should be avoided in these strains to prevent treatment failure.

Another salient finding was that the genotyped XDR K. pneumoniae strains (n = 2) carried the bla_SHV−1 gene. This gene is reported only in resistant K. pneumoniae strains and encodes intrinsic resistance to β-lactam antibiotics (27, 28). Thus, it can be inferred that β-lactam antibiotics should be used with caution against any resistant K. pneumoniae strains.

Earlier, Tayyaba et al. (29) had characterised MDR and XDR clinical isolates of A. baumannii using WGS, which revealed the presence of adeL and adeK genes specific to A. baumannii strains, encoding efflux pumps of the adelJK operon. These findings were consistent with our study and were also supported by previous work of Ghafari et al. (30), who also examined the presence of adelLJK efflux pump genes in clinical isolates of A. baumannii isolated from wound and urine samples.

Furthermore, AMR genotyping revealed a complex resistome, with bla_OXA, bla_TEM−1, and bla_CTX−M−15 predominating among Gram-negative isolates, along with functional AMR genes conferring resistance to tetracyclines (tetA/tetB) and fluoroquinolones (qnrB, gyrA, parC), as well as multiple active efflux pumps. The identification of additional resistance determinants (e.g., fosA, sul1/sul2, aac, cat) against antibiotics that were not included in the phenotypic screening panel suggests a broader resistance potential beyond routine antibiogram testing.

A key strength of this study lies in the strain-wise comparative analysis of functional resistance genes compared with the standard reference genomes of clinical isolates (Figure 3). This matrix-based evaluation effectively demonstrates significant intra-species heterogeneity, indicating ongoing adaptive evolution through horizontal gene transfer under antibiotic pressure (31). While several studies have performed comparative genomic and resistome analyses of ESKAPE pathogens (32, 33), the present work differs in important aspects, as it integrates phenotypic susceptibility data, intrinsic AMR gene identification, and functional resistome profiling of the clinical strains of XDR-ESKAPE group, isolated from blood and BALF samples of sepsis patients. Comprehensive genomic information can be used to design personalised treatment regimen, which can be curated based on the presence of functional AMR genes in the specific causal organism (34). This exhaustive molecular-level examination has created a genetic database of sepsis-causing XDR-ESKAPE organisms, which helped to formulate pathogen-specific antibiotic combinations and validate them in vitro. The work also provides a useful framework to create the blueprint for handling XDR-ESKAPE in sepsis samples so as to overcome resistance and influence potential antibiotic stewardship.

5 Conclusion

The increasing antibiotic resistance in sepsis constitutes a major global health challenge, requiring an in-depth evaluation of causal organisms and critical reassessment of standard monotherapies as the exclusive strategy for empiric treatment. The integration of techniques such as WGS can help characterise the sepsis-causing pathogen(s), facilitate analysis of emerging antibiotic resistance patterns, and greatly strengthen AMR surveillance. The phenotypic/genomic analysis of the XDR-ESKAPE in this study indicates that the use of antibiotics ciprofloxacin, ofloxacin, levofloxacin, imipenem, and ampicillin/sulbactam as monotherapy should be done very diligently for sepsis patients, as the genes responsible for resistance to these antibiotics were intrinsic to all strains of XDR-ESKAPE.

Also, combining the knowledge of antibiotic resistance patterns and the functional AMR gene profile in the XDR-ESKAPE organisms, a genomic databank of the antibiotic-resistant sepsis-causing pathogen(s) has been created. With this serving as reference, tailored pathogen-specific antibiotic combinations can be formulated to override the resistance mechanisms adopted by the ESKAPE group in sepsis.

As the global burden of ICU-related sepsis intensifies, this provides a model plan that could be adopted for future multicentre clinical trials aimed at validating this molecular-guided strategy in diverse populations.

6 Limitations

The present study has several limitations that should be acknowledged. First, the study was conducted at a single tertiary care centre, which may limit the generalizability of the findings with respect to other healthcare settings of different patient demographics, clinical practices, and microbial profiles. Second, the relatively limited sample size may have restricted the resistance patterns of the isolated XDR-ESKAPE in reference to varied geographical regions. Third, whole-genome sequencing was performed on only 12 ESKAPE isolates due to resource limitations that possibly confines the representation of genomic diversity and resistance determinants of the sepsis cohort. Finally, although promising antibiotic combinations were identified for potentiating the ineffective resistant agents, at present these combinations were evaluated solely through in vitro testing. They lack clinical validation in terms of patient outcomes and therapeutic efficacy.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/femer.2026.1884260/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Gabriel Wardi, University of California, San Diego, United States

Reviewed by: Michel Manika Muteya, University of Lubumbashi, Democratic Republic of Congo

Ariful Basher, Infectious Diseases Hospital, Bangladesh