Stool-based genomic surveillance identifies post-engraftment expansion of multidrug-resistant pathogens in haematopoietic stem cell transplant patients in India

Stool-based genomic surveillance identifies post-engraftment expansion of multidrug-resistant pathogens in haematopoietic stem cell transplant patients in India | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Stool-based genomic surveillance identifies post-engraftment expansion of multidrug-resistant pathogens in haematopoietic stem cell transplant patients in India Archana Madhav, Jobin John Jacob, Sushil Selvarajan, Sanika Kulkarni, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9289820/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Patients undergoing haematopoietic stem cell transplantation (HSCT) are highly susceptible to bloodstream infections (BSIs) caused by antimicrobial resistant (AMR) pathogens, yet the temporal dynamics and clinical relevance of intestinal AMR reservoirs remain poorly defined, particularly in high-burden settings. Methods We conducted prospective longitudinal surveillance of 81 HSCT recipients in India using targeted enrichment-based stool metagenomics, coupled with whole genome sequencing of bloodstream isolates and strain-level tracking. Results Across 252 samples, we identified a restructuring of the gut-associated resistome, characterised by depletion during conditioning and a marked post-engraftment expansion of non-efflux AMR determinants, including plasmid-borne carbapenemase and ESBL genes ( bla NDM, bla OXA). This expansion defined a previously underappreciated window of vulnerability during early immune recovery. Over 50% of patients harboured multidrug-resistant organisms, and carriage of bla NDM was associated with increased odds of subsequent infection. Strain-resolved analyses provided genomic evidence linking gut colonisation to bloodstream infection in a subset of cases, demonstrating the feasibility of non-invasive prediction of invasive disease. Patients with AMR-BSIs experienced high mortality (> 40%). Conclusions These findings establish that dynamic changes in the gut resistome following HSCT can identify periods of heightened infection risk and provide a foundation for predictive, genomics-guided surveillance and antimicrobial stewardship strategies in high-burden settings. Biological sciences/Computational biology and bioinformatics Health sciences/Diseases Biological sciences/Genetics Biological sciences/Microbiology Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Between 15–65% of all haematopoietic stem cell transplantation (HSCT) patients develop bacterial bloodstream infections (BSI), two-thirds of which are associated with mucosal barrier injury (MBI) ( 1 ). Enterobacteriaceae are the most common aetiologic agents of Gram-negative bacteraemia in HSCT patients ( 2 ), accounting for ~ 75% of culture-positive BSIs post-HSCT in India ( 3 ). Risk of Enterobacteriaceae BSIs is amplified by their potential to accumulate antimicrobial resistance (AMR) determinants, which may be exacerbated by stress induced by intestinal inflammation ( 4 ), with the gut acting as a reservoir of AMR genes ( 5 ). Sardzikova et al. reported that multi-drug resistant (MDR) E. coli , Klebsiella pneumoniae , Enterococcus faecium and Staphylococcus aureus are leading contributors to the resistome in the HSCT gut ( 6 ). Low-middle income countries (LMICs) consistently report a higher prevalence of AMR in almost all bacteria, yet syndrome-specific data (e.g., AMR in patients with BSIs post-HSCT) are scarce ( 7 ). The cocktail of immunosuppressants and antimicrobials administered post-HSCT reduce the sensitivity of culture-based diagnostics of BSI, thus limiting the application of antimicrobial susceptibility testing (AST) ( 8 ). The growing challenge posed by AMR in LMICs ( 9 ), coupled with the limited HSCT patient data in this region, underscores the need for further investigation, where genomic approaches could offer new insights and solutions. Here, we enrolled HSCT patients at a tertiary care centre in India and utilised plate-sweep enrichment on longitudinally collected stool samples. We targeted priority AMR pathogens frequently associated with BSIs post-HSCT and their associated resistome trajectories. We additionally leveraged enhanced metagenome-assembled genome (MAG) recovery from enriched samples and conducted whole-genome sequencing (WGS) of bloodstream isolates from BSI episodes to detect organisms co-present in the gut and bloodstream of patients with culture-positive BSI. Lastly, we derived plasmid-host associations for patients with culture-positive BSIs post-HSCT using Hi-C on the sweep-enriched fraction of stool samples and tracked sweep-enriched strains within and between participants in our cohort. Methods Study design We conducted a prospective longitudinal study enrolling 81 participants undergoing allogeneic HSCT between May and December 2022 at the Department of Haematology, Christian Medical College (CMC), Vellore, India. The study was a collaboration between the Cambridge Institute for Therapeutic Immunology and Infectious Disease (CITIID) and the Clinical Microbiology and Clinical Haematology departments at CMC Vellore. This study was granted ethical approval by the Institutional Review Board (IRB), Research and Ethics committee at CMC Vellore in accordance with the Declaration of Helsinki and the Indian Council of Medical Research National Ethical Guidelines for Biomedical and Health Research involving Human Participants (IRB Min No. 14605 dated 27.04.2022). Participant eligibility criteria and clinical data management Patients admitted to the Department of Haematology at CMC Hospital were eligible if scheduled to undergo an allogeneic stem cell transplant and provided informed consent to participate. Informed consent forms were written in English and translated into Hindi and Tamil. A clinical case report form recorded clinical metadata including age, sex, diagnosis, donor age, donor sex, HLA match, type of transplant, preparative regimen, post-HSCT infection episodes, episodes of Clostridioides difficile colitis, graft-versus-host disease (GvHD) and mucositis, date of absolute neutrophil count (ANC) recovery, total days of hospital stay, and outcome. Additional clinical information was obtained from the clinicians or nurses in-charge, accessed through the clinical workstation, or through discharge/death summaries. Sensitive clinical information was accessed through secure hospital servers and shared only with key personnel. Stool sample collection and processing Stool samples were collected in sterile stool collection containers and transferred to the Department of Clinical Microbiology at 4ºC. Once received, samples were labelled and processed within 72 hours. Blood samples were collected for culture and AST if a BSI was suspected by the clinician in-charge, for example, if the patient developed a fever post-HSCT. Stool samples were homogenised in a 1:1 ratio with PBS and 100x dilutions were plated on MacConkey agar plates without crystal violet to grow Gram-negative bacteria and Gram-positive bacteria belonging to the Staphylococcus and Enterococcus genera. Following overnight incubation, a complete sweep of the plate was collected by resuspending any grown colonies in sterile PBS. DNA extraction was done on all plate sweep aspirates using the Wizard® Genomic DNA purification kit by Promega per the manufacturer’s protocol. A subset of plate sweep aspirates (selected retrospectively for patients with post-HSCT culture positive BSI) additionally underwent formalin inactivation for Hi-C. Detailed protocols for stool sample processing, serial dilutions, plate sweeps, DNA extraction and formalin inactivation can be found in supplementary methods. Blood sample collection and processing Blood culture bottles were incubated in the BacT/ALERT® automated Blood Culture system. Positive samples were passaged either on blood agar or selective media chosen according to the pathogen detected to obtain pure isolates for WGS. Isolates were incubated overnight in nutrient broth at 37ºC, and a 1 mL aliquot of the overnight culture was used for DNA extraction and WGS. Genome sequencing DNA concentrations were quantified on a Qubit™ 2.0 Fluorometer, and shipped to Eurofins Genomics for standard library preparation with equimolar pooling and sequencing. Samples were sequenced on an Illumina NovaSeq 6000 S4 PE150 instrument, with a minimum of 10 million paired-end reads generated for whole metagenomic sequencing (WMS) of enriched stool samples and 100x coverage for whole genome sequencing (WGS) for bloodstream isolates. The sequencing reads for both stool and blood samples from the study can be found under BioProject accession PRJNA1072756. Two-hundred and ninety-eight (298) enriched stool samples were sent for WMS. Twenty samples contained insufficient DNA to generate QC-checked libraries (< 1nM/L of DNA). After removing samples with incomplete metadata or belonging to patients excluded based on patient eligibility criteria, a total of 252 samples from 81 patients were included in the study. For Hi-C, 13 formalin-inactivated enriched stool plate sweeps were sent to Phase Genomics in Seattle, USA for Hi-C library prep and sequencing according to the ProxiMeta ™ Hi-C Kit Protocol v4.5. Contigs assembled from enriched stool metagenomic read data were sent to the bioinformatics team at Phase Genomics to overlay Hi-C connections. Two stool samples (P14AT0065 and P51AT1247) did not pass QC for Hi-C. Quality control and assembly of sequencing reads Raw sequencing reads from the enriched stool samples were run through the nf-core/mag pipeline v2.3.0 ( 10 ) for trimming, adapter removal, and quality control with FastQC. Enriched stool metagenomic read assembly was performed using SPAdes v3.15.4, and all assemblies underwent QC using QUAST ( 11 ). We used Metawrap v1.3.0 ( 12 ) to produce consensus bins from three binning pipelines, MetaBAT2 v2.12.1 ( 13 ), Maxbin2 v2.2.7 ( 14 ) and CONCOCT v1.1.0 ( 15 ). The threshold for consensus bins was set to a minimum of 70% completeness and < 5% contamination to extract bins of high purity. After extracting consensus bins, each bin was passed through GTDB-TK ( 16 ) for species identification. Taxonomic assignment and microbial diversity Taxonomic assignment was performed directly on enriched stool plate sweep reads using Kraken v2.1.2 ( 17 ), followed by Bayesian re-estimation at genus level using Bracken v2.7.0 ( 18 ). Residual human OTUs were removed with the “extract_kraken_reads.py” script. Bracken outputs were combined using the “combine_kraken_reports.py” script and converted into a .biom format using kraken-biom v1.2.0 from the KrakenTools software suite ( 19 ). OTU tables were imported into RStudio v4.2.3 and visualisations for taxonomic composition of the enriched fraction of stool samples was performed using metacoder v0.3.6 ( 20 ). Beta diversity was calculated using the Bray-Curtis dissimilarity index with the vegdist function from the vegan package v2.5-7 ( 21 ). A pairwise adonis test was performed using the pairwiseAdonis v0.4.1 ( 22 ) package on R. AMR gene detection and antimicrobial sensitivity testing AMR genes were detected from bloodstream isolates and enriched stool sequencing data using ABRIcate v1.0.1 with the CARD database as reference with default parameters. Antimicrobial drug classes were reclassified as ‘multidrug’ if the AMR gene detected conferred resistance to \(\:\ge\:\) 3 antimicrobial classes. Phenotypic AST was performed on bloodstream isolates and stool colony picks using 6mm disks of Whatmann No. 1 filter paper saturated with antimicrobial solutions. Stool colony picks were tested against 8 antimicrobials for Gram-negative organisms and 4 antimicrobials for Gram-positive organisms. A complete description of drug-brug AST combinations for stool colony picks is provided in supplementary methods. InStrain analysis and microbial network InStrain v1.8.0 ( 23 ) was used in this study for two objectives 1) to identify episodes where bloodstream isolates from patients with culture positive BSI post-HSCT were also present in the gut, suggesting translocation, and 2) to detect strains from priority AMR organisms shared across multiple enriched stool samples. Correspondingly, we used two different ANI thresholds when attempting to detect isolates across samples. For detecting possible translocation events, we used a 99.9% population-level ANI to account for variation introduced by differences in sample processing and sequencing between blood and stool samples and a 50% benchmarked breadth threshold as recommended in the inStrain pipeline documentation to minimise risk of false positives. For detecting shared isolates across stool samples, we used thresholds of 99.999% population-level ANI and at least 25% ‘genome compared’ between samples. Details of inStrain pipeline configuration and output analyses can be found in supplementary methods. The microbial network depicting strains from targeted AMR organisms shared across multiple enriched stool samples was created using the igraph v1.5.1 ( 24 ), tidygraph v1.2.3 ( 25 ) and ggraph v2.1.0 ( 26 ) packages on RStudio. Each node ( n = 98) represents an enriched stool sample and edges between nodes denote connections ( n = 174), i.e., bacterial strains shared between samples. The number of connections to each sample was used to produce a network plot using ggraph under the ‘kk’ layout. Results Cohort summary and outcomes A summary of the baseline clinical characteristics of the 81 patients in the allogeneic HSCT cohort is shown in Table 1 . Overall, 32/81 patients developed a clinically significant bacterial infection (38 episodes total) post-HSCT. Clostridioides difficile colitis was the most common infection post-HSCT ( n = 11), followed by Enterobacteriaceae ( n = 7) and coagulase negative Staphylococcus (CONS) ( n = 7, including 2 episodes of central-line associated BSI). Excluding catheter and central-line associated BSI (n = 3), 17 patients experienced culture-positive bacterial BSIs post-HSCT (20 episodes in total). There were 10 episodes of culture-positive BSI prior to HSCT among seven patients (resolved before transplantation). Table 1 Baseline characteristics of the HSCT study cohort Baseline characteristics n (%) Age (years) < 18 38 (46.9%) 18–30 13 (16.0%) 31–40 13 (16.0%) 41–50 11 (13.6%) 51–60 6 (7.4%) Sex (% male) 53 (65.4%) Diagnosis Malignant 39 (48.1% ) Acute lymphoblastic leukaemia 13 (33.3%) Acute myeloid leukaemia 9 (23.1%) Myelodysplastic syndrome 8 (20.5%) Chronic myeloid leukaemia 5 (12.8%) Hodgkin's lymphoma 2 (5.1%) Polycythemia vera 1 (2.6%) NK T-cell lymphoma 1 (2.6%) Non-malignant 42 (51.9%) Thalassemia major 17 (40.5%) Severe aplastic anaemia 15 (35.7%) Fanconi anaemia 4 (9.5%) Sickle cell anaemia 1 (2.4%) Wiskott Aldrich syndrome 1 (2.4%) Mucopolysaccharidosis 1 (2.4%) Chronic granulomatous disease 1 (2.4%) Severe congenital neutropenia 1 (2.4%) Paroxysmal nocturnal haemoglobinuria 1 (2.4%) Type of allogeneic transplant Fully matched 51 (63.0%) Related donor 40 (49.4%) Unrelated donor 11 (13.6%) Haploidentical 30 (37.0%) Conditioning regimen Myeloablative 58 (71.6%) Reduced intensity 23 (28.4%) Outcome Discharged 66 (81.5%) Died 15 (18.5%) The median time to culture-positive BSI was 12 days post-HSCT. We recovered 19 bacterial BSI isolates on culture plates, which were subjected to WGS (7/10 pre- and 12/20 post-HSCT). Gram-negative bacilli (GNB) caused nearly two-thirds of post-transplant BSI episodes (13/20). The mortality rate among patients who developed culture-positive BSI post-HSCT was 41.2% (7/17), whereas the mortality rate among patients who developed any infectious complication post-HSCT was 37.5% (12/32). Supplementary data (SD) 1 contains a summary of all clinically significant organisms detected post-HSCT in our cohort. Fifty-seven patients had mucositis (median grade 3). Out of 81 patients, 21 experienced GvHD, of whom 15 had gastrointestinal (GI)-GvHD. Over half of all patients who had GI-GvHD (8/15) also had episodes of Clostridioides difficile infection. Patients 12, 18 and 26 had haploidentical HSCTs, developed overall GvHD at grade 4, and all three had Clostridioides difficile infections. Patients 12 and 18 succumbed to complications they developed 51- and 77-days post-admission respectively, but patient 26 recovered and was discharged 93 days after admission. Plate-sweep enrichment enables high-resolution recovery of priority AMR pathogens A median of three stool samples was collected from each patient (range 1–8), with at least three stool samples collected for 61/81 patients. Stool samples (n = 252) were categorised into five groups based on the timepoint relative to HSCT: pre-conditioning (before day − 8; n = 12), conditioning week (day − 8 to day 0; n = 78), pre-engraftment (day + 1 to day + 15; n = 89), early post-engraftment (day + 16 to day + 100; n = 67) and late post-engraftment (day + 101 to day + 200; n = 6). Figure 1 outlines the bacterial taxa that were enriched in the MacConkey plate sweeps. We specifically targeted Gram-negative organisms (i.e., E. coli , Klebsiella pneumoniae , Acinetobacter baumannii and Pseudomonas aeruginosa ) but the omission of crystal violet permitted the growth of Staphylococcus spp. and Enterococcus spp. Therefore, apart from Clostridioides difficile , our MacConkey plate sweep enrichment approach supported the detection of all clinically significant organisms that emerged in our HSCT cohort (SD1). We intentionally selected for priority AMR pathogens in the HSCT cohort with this approach, as described by the focused colour dispersion within specific branches of Pseudomonadota (Proteobacteria; Fig. 1 A) and Bacillota (Firmicutes; Fig. 1 B). The 252 stool samples generated between 1.8 to 33 million reads per sample, with nearly 3.5 billion individual operational taxonomic units (OTUs) spanning 4,272 taxonomic levels. This dataset was rarefied at the 10th percentile, representing approximately half of the original dataset with over 1.6 billion OTUs. Over three-quarters (77.6%; n = 1,282,928,757) of the OTUs belonged to the Enterobacteriaceae family in the Pseudomonadota (Proteobacteria) phylum. At genus level, this family mainly consisted of Escherichia (73.4%), Klebsiella (21.8%), Shigella (1.54%), Citrobacter (1.38%) and Enterobacter (1.34%). In the Bacillota (Firmicutes) phylum ( n = 217,106,676), Lactobacillales and Bacillales orders comprised 89.6% and 10.4% of OTUs in the phylum, respectively. The Lactobacillales order was dominated by Enterococcus spp. (99.8%), and the genus Staphylococcus was the most abundant (99.6%) in the Bacillales order. The metacoder plot hence depicts a selectively enriched subcommunity of organisms from stool samples that have previously been identified as leading contributors to AMR and BSI among HSCT patients and notably does not capture total gut microbiome composition. Taxonomic diversity of enriched stool from HSCT patients A Shannon alpha diversity analysis of enriched stool samples at species-level, as expected, did not show any significant difference (one-way ANOVA) when grouped by timepoints relative to HSCT (SD2), suggesting even enrichment of the intended subcommunity across samples. Beta diversity was calculated to examine the population level differences in the enriched subcommunities relative to the transplant timeline using the Bray-Curtis dissimilarity index (Fig. 2 A). The plot shows clustering of conditioning week samples (orange). A pairwise adonis test was used to probe grouped timepoint beta diversities in a pairwise manner (*** p < 0.001; SD3). These data depicted a notable restructure among the enriched subcommunity during the conditioning week when compared to samples from the pre-engraftment (R² 0.156), early- (R² 0.105), and late post-engraftment timepoints (R² 0.161); the genera contributing to these differences are shown in SD4. No significant associations were observed when assessing beta diversity of the enriched subcommunity in relation to transplant outcomes or GvHD. However, patients who developed acute GvHD post-HSCT demonstrated significant log 2 fold-increases in the Enterococcus (*** p < 0.001) and Morganella (*** p < 0.001) genera (Fig. 2 B). Temporal dynamics of AMR in target organisms ABRIcate was used to detect AMR genes among the enriched taxa from the HSCT cohort’s stool samples using the CARD database. 16,903 AMR gene hits were detected among the 252 enriched stool samples, with efflux-mediating genes representing 65.8% ( n = 11,118) of AMR genes detected within the collection. The median number of AMR genes detected among enriched taxa in the pre-conditioning phase ( n = 12) and conditioning week ( n = 78) were 62 and 68 AMR genes per sample, respectively, compared to a median of 55 AMR genes detected per enriched sample during pre-engraftment ( n = 89). Correspondingly, a median Shannon alpha diversity index of 3.23 was observed for AMR genes during both the pre-conditioning and conditioning phases (Fig. 3 A). At the pre-engraftment timepoint, the Shannon index dropped to 3.06. These decreases in the abundance and diversity of AMR genes between the conditioning week and pre-engraftment timepoints were both significant at *** p < 0.001 with a pairwise t-test with Bonferroni correction (SD5). At early post-engraftment, there was a significant increase in the abundance (** p < 0.01) and diversity (*** p < 0.001) of AMR genes compared to the pre-engraftment timepoint among the enriched taxa. The median number of AMR genes detected per enriched sample at early post-engraftment was 69 (n = 67; alpha diversity = 3.32) and 96 at late post-engraftment (n = 6; alpha diversity = 3.46). The abundance of non-efflux AMR genes within the enriched subcommunity was significantly higher at early (** p < 0.01) and late (* p < 0.05) post-engraftment compared to pre-engraftment, suggesting that organisms within the enriched subcommunity carrying non-efflux AMR genes remained in the HSCT gut post-transplantation. This observation suggests preferential selection through robust tolerance and survival mechanisms among these organisms but should be interpreted with caution since only six samples were collected at late post-engraftment. The abundance of five MDR-conferring ESBL/carbapenemase genes ( n = 695) – blaOXA , blaNDM , blaCTX-M , blaSHV and blaTEM were plotted longitudinally against the HSCT timeline in Fig. 3 C. At least 1/5 MDR genes were detected in 216/252 enriched stool samples. Overall, we observed a significant successive increase in the abundance of MDR genes when comparing enriched samples collected during the early post engraftment timepoint to MDR genes detected in enriched samples during the preceding conditioning week (*** p < 0.001) and the pre-engraftment (** p < 0.01) timepoints. Notably, the blaOXA and blaNDM carbapenemase genes exhibited progressive increases in their mean abundance per sample between the conditioning week (0.60 and 0.23, respectively) and late post-engraftment (1.40 and 0.8, respectively). In multivariable logistic regression, most AMR genes were not associated with infection, except for a ~5x increased odds of infection among patients who carried the blaNDM gene during the conditioning week ( p = 0.042; SD6). By using a Hi-C approach on the sweep-enriched fractions of stool samples obtained from patients with culture positive BSI post-HSCT, we found that blaOXA and blaNDM genes often co-localised on plasmids harboured by E. coli and Klebsiella pneumoniae . Such plasmids were detected in half of all patients who developed culture positive BSIs; however, with non-identical AMR gene compositions (Fig. 3 D). These plasmids also carried genes inducing resistance against tetracyclines, trimethoprim, sulphonamides, and macrolides. Among the 11 samples which underwent Hi-C, 70.3% of AMR genes detected were of plasmid origin; the plasmid-associated AMR gene to host summary is provided in SD7. AST data for select enteric MDR bacteria ( E. coli , Klebsiella pneumoniae and Enterococcus faecium ) isolated from stool colony picks ( n = 90) alongside a subset of AMR genes from enriched stool samples are depicted in SD8. MDR isolates were detected in 57 samples from 41 patients, signifying that > 50% of patients had at least one phenotypically MDR organism isolated from their stool during the HSCT timeline. Spatiotemporal isolate tracking reveals persistence and possible transmission Plate-sweep metagenomics offers a factor-increase in the resolution of MAGs recovered from stool samples. This approach enabled us to probe microdiversity within enriched taxa to identify strains shared across multiple patients within the cohort, and map bloodstream isolates from clinical infection episodes which underwent WGS to enriched stool samples. Figure 4 shows a microbial network which depicts (at strain level) organisms from the selective subcommunity that were detected across multiple stool samples. To detect the same isolate among multiple participants, or across multiple timepoints within the same participant, we used a stringent average nucleotide identity (ANI) threshold of 99.999% at the population-level (popANI), where at least 25% of the enriched stool reads were compared against the reference consensus MAG bins extracted from the plate sweep sequencing data. The resulting microbial network consisted of 98 nodes (samples) and 174 edges (strains); the degree annotation represents the number of edges (i.e., strains) that connect to each node (Fig. 4 ). The network depicts clusters where sequencing reads from the enriched stool samples mapped with \(\:\ge\:\) 99.999% popANI to reconstructed MAG reference bins, indicating highly similar isolates detected across multiple samples. There were several episodes (participants 3, 11, 46, 54, 57, and 61) where isolates could be linked longitudinally within the same participant, and one participant (participant 36) who had MAGs from Escherichia coli and Providencia alcalifaciens (> 85% complete, <5% contaminated) that could be linked across timepoints. Transmission links where two edges were shared between samples indicate episodes where reads from both enriched stool samples mapped to reconstructed MAGs from both samples at the \(\:\ge\:\) 99.999% ANI and \(\:\ge\:\) 25% ‘genome compared’ thresholds. Cases where these bidirectional relationships appeared across multiple participants likely indicate localised transmission clusters but would need confirmation with environmental surveillance data. Genomic evidence links gut colonisation to bloodstream infections in select cases The cultured organism from the blood samples was matched to an organism present in the selectively enriched stool fraction in five patients at strain-level – patient 2 ( E. coli ), 5 ( Enterococcus faecium ), 14 ( Klebsiella pneumoniae ), 44 ( Pseudomonas aeruginosa ) and 55 ( Klebsiella pneumoniae ). All five patients presented with malignant indications for HSCT, underwent myeloablative conditioning, and developed mucositis (median grade 3). Patients 2, 5, 14 and 44 had episodes of culture-positive BSI post-HSCT, whereas patient 55 had culture-positive BSI 45 days pre-HSCT. In patients 2, 5, and 14, the blood culture isolate was matched to enriched stool collected prior to BSI, suggesting temporally plausible gut-blood translocation, and in patients 44 and 55, the BSI isolate was matched to enriched stool collected 16 and 100 days after BSI, i.e., temporally ambiguous. In patient 55, the blood culture and AST report identified the organism as carbapenem-resistant Klebsiella pneumoniae. After HSCT, patient 55 developed febrile neutropenia with septic shock and grade 3 mucositis, but empirical treatment with meropenem/colistin/teicoplanin likely contributed to subsequent sterile blood cultures. While the clinical infection in patient 55 occurred 45 days before HSCT and was resolved prior to transplantation, the post-HSCT detection of the BSI-causing Klebsiella pneumoniae strain in the enriched stool sweep suggests that the organism persisted within the host in the gut. In 15.8% (3/19) of sequenced BSI cases, we obtained direct genomic evidence of prior gut colonisation by the causative pathogen. All five BSI isolates that were co-detected in the gut were genotypically MDR and four were phenotypically MDR (SD9); patient 14 died before AST could be performed. A clinical summary of the patients with BSI isolates mapping to a corresponding enriched stool bin is provided in Table 2 . Table 2 Summary of bloodstream isolates from BSI episodes which mapped to the corresponding participant’s gut microbiome sample. Bloodstream isolates underwent WGS where reads mapped to reconstructed MAGs (> 97% completeness, 99.9% ANI across > 50% breadth of the reconstructed MAG. bloodstream_isolate genome bin_completeness (%) bin_contamination (%) gtdbtk_species_classification breadth conANI popANI P02AT005BP2 P02BT0031_bin.2 95.37 0.04 Escherichia coli 0.63 0.9999 0.9999 P05AT014BA1 P05AT0063_bin.2 99.25 0.19 Enterococcus faecium 0.64 0.9997 0.9998 P14AT043BA1 P14BT0061_bin.1 97.00 0.81 Klebsiella pneumoniae 0.77 0.9998 0.9998 P44AT012BA1 P44AT0283_bin.2 99.68 0.62 Pseudomonas aeruginosa 0.99 0.9999 0.9999 P55BT045BA1 P55AT0554_bin.2 98.80 1.14 Klebsiella pneumoniae 0.54 0.9999 0.9999 Discussion HSCT patients in LMICs are highly vulnerable to post-transplant BSIs caused by AMR pathogens. This study builds on a previous investigation by Korula et al. ( 27 ) in CMC, where bacterial translocation was inferred by matching organisms and AST results between isolates recovered from blood and stool cultures. Here, we used genomics to conduct prospective longitudinal surveillance of stool samples collected from 81 HSCT patients at the same hospital in India, by deploying plate-sweep enrichment to select for targeted priority pathogens, alongside WGS of bloodstream isolates from BSI episodes. Our findings revealed three key areas for further research and potential clinical impact. First, MacConkey-enriched microbial populations in the HSCT gut were significantly modulated by the pre-transplant conditioning regimen, and their associated resistomes were depleted at the pre-engraftment phase. Wong et al. reported analogous longitudinal shifts in total faecal microbiome beta-diversities, using 16S data from an Asian cohort undergoing autologous HSCT ( 28 ), which recovered to preconditioning levels within 6-months post-HSCT (beyond our timeframe for follow-up). However, the resistomes in our cohort were promptly restored at early post-engraftment and shifted towards a sustained and significant increase in the abundance of non-efflux AMR genes, including MDR genes, into the late post-engraftment phase. In a longitudinal study with eight paediatric patients in Italy, D’Amico et al. had similar observations with the total gut resistome ( 29 ), reporting an AMR gene profile that had diversified with MDR genes in addition to consolidating the resistome present prior to HSCT. This temporal signature suggests that the early post-engraftment period, when patients are typically recovering and antibiotics administered to counter febrile neutropenia are being de-escalated, paradoxically represents a window of increased AMR risk, which warrants enhanced surveillance to inform antimicrobial stewardship efforts. Here we generated direct genomic evidence to show that gut colonisation precedes invasive infection in 15.8% of BSI cases in our cohort, establishing proof-of-concept that enrichment-based genomics can identify patients who subsequently developed invasive disease. Blood culture positivity is lower among HSCT patients, since the antimicrobials administered to counter febrile neutropenia during pre-engraftment further reduce the low sensitivity of blood cultures ( 30 ). All five patients who had isolates matched between blood and enriched stool samples also showed evidence of enteric mucositis, which has been linked with infections by Gram negative Enterobacteriaceae , Pseudomonas aeruginosa and Gram-positive Enterococci ( 31 ). Additionally, the enrichment-based strategy offers an alternative mechanism to monitor localised strain transmission of clinically significant pathogens, especially ESBL Enterobacterales ( 32 ), demonstrated by our ability to track isolates with 99.999% ANI longitudinally within and between patients across timepoints. The fact that ~ 74% of BSI cases could not be linked to the gut highlights the complexity of infection pathogenesis in HSCT recipients and that multiple acquisition routes (nosocomial, respiratory, or gut-associated below detection thresholds) likely contribute. We demonstrated that > 50% of HSCT recipients carried MDR organisms during the transplant timeline, often harbouring ESBL and carbapenemase genes, which co-occur on diverse plasmids across E. coli and Klebsiella pneumoniae species. Enterobacteriaceae bloom under conditions of intestinal inflammation ( 4 ), which also exacerbates horizontal gene transfer and nosocomial spread in this high-risk population. While WMS hinders the association of AMR genes back to their host species, using plate-sweep enrichment reduces the pool of potential gene hosts compared to WMS, and permits high-resolution screening of targeted bacterial populations compared to isolate-based approaches. Ghosh et al. reported that the high prevalence of MDR in an Indian HSCT cohort was primarily associated with the blaNDM and blaOXA - 48-like carbapenemase genes ( 33 ), but only with pre-transplant surveillance samples. With the longitudinal sampling strategy, we found multiple blaNDM and blaOXA genes among the enriched taxa in our cohort, and this abundance significantly increased in the early post-engraftment timepoint when compared to the two preceding timepoints. The microbial clusters we identified suggest possible transmission clusters although environmental surveillance data are required for confirmation. Methodologically, this study establishes that plate-sweep enrichment offers a practical middle ground between traditional culture, which relies on single colonies and does not provide population or strain-level information, and deep WMS which may be too costly to be conducted at scale to inform clinical decision making. Our data suggest that patients with expanding post-engraftment ESBL/carbapenemase gene abundance could be candidates for modified empirical therapy, pre-emptive decolonisation or enrolment in microbiome restoration trials. However, whether resistome-guided interventions improve outcomes compared to standard care remains an open question requiring randomised controlled trials. We did not consider the clinical heterogeneity of HSCT patients. Future investigations could focus on more specific cohorts based on their age, indications for HSCT or a defined outcome, e.g., GI-GvHD. Secondly, selective enrichment can underrepresent slow-growing, fastidious, or anaerobic bacterial populations and by extension, their resistomes. However, our enrichment-based approach was selected to accommodate routine microbiological testing prior to research. Prolonged storage can skew the metagenomic composition of the microbiota in the stool, allowing populations of facultative anaerobes to flourish upon exposure to oxygen if stool collection containers were opened repeatedly, whereas the proportion of obligate anaerobes would stagnate. While our approach limits comparisons with WMS data generated from other HSCT cohorts and detection of rare gut species which could be clinically significant, the MacConkey plate enrichment helped standardise over inconsistencies in sample handling and storage, by selecting key pathogenic members of the microbiota previously identified as important contributors to the burden of AMR infections in HSCT patients. While the enrichment strategy outperforms WMS from a neat sample in capturing intraspecific or population-level heterogeneity for the selected organisms, binning algorithms capture the most dominant STs ( 12 ). Nonetheless, we were able to account for the population-level heterogeneity across samples because of inStrain’s unique ‘microdiversity-aware’ metrics of conANI and popANI ( 23 ). Therefore, although we can report bacterial strains co-detected across samples and sample types, we cannot infer transmission source and direction for the remaining ~ 74% of BSI episodes without denser temporal sampling of HSCT patients, more robust spatial tracking, and environmental surveillance. In conclusion, we have established the first genomic baseline for AMR dynamics among targeted priority pathogens in HSCT patients in South Asia and provide proof-of-concept that gut reservoirs can be linked to invasive infections. While questions remain about the relative contribution of gut translocation versus other BSI sources, the post-engraftment resistome expansion we observed represents a potentially modifiable risk factor and a clear target for future studies. As HSCT programs expand in LMIC regions with high AMR burden, such surveillance approaches which leverage non-invasive sampling and enrichment-based sequencing will become essential tools for protecting vulnerable transplant populations. Declarations Funding and acknowledgements This work was supported by a Wellcome Senior Research Fellowship (215515/Z/19/Z) to Stephen Baker and a grant from the Bill and Melinda Gates Foundation (OPP1159351). The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. We would like to thank Agila Kumari Pragasam, Chaitra Shankar, and Ellen Higginson for helpful discussions on the study design. We are also grateful to the media preparation team in the department of Clinical Microbiology at CMC Vellore, and Plamena Naydenova for her assistance with quantifying DNA concentrations. We would also like to thank Jolynne Mokaya for critical feedback on the manuscript. Finally, we would like to acknowledge the clinical teams for their assistance with sample collection and thank the patients and their families for their participation and support. Conflicts of Interest The authors report there are no competing interests to declare. Author Contribution Conceptualisation and design: AM, SB, BV, SS, BGCollection of samples and clinical data: SS, SD, BGProcessing of samples: AM, SK, YM, DM, PS, KMAnalysis and interpretation of data: AM, JJJ, SBDrafting initial paper: AM, JJJ, SBRevising the final paper: AM, JJJ, SS, SK, SD, YM, DM, PS, KM, BG, BV, SB Data Availability The authors confirm that the data supporting the findings of this study are available in the article and its supplementary materials. The genomic sequencing reads have been deposited in the NCBI Sequence Read Archive under BioProject accession PRJNA1072756. References Heston SM, Young RR, Hong H, Akinboyo IC, Tanaka JS, Martin PL, et al. Microbiology of Bloodstream Infections in Children After Hematopoietic Stem Cell Transplantation: A Single-Center Experience Over Two Decades (1997–2017). Open Forum Infect Dis. 2020;7(11):ofaa465. doi: 10.1093/ofid/ofaa465 Pouch SM, Satlin MJ. Carbapenem-resistant Enterobacteriaceae in special populations: Solid organ transplant recipients, stem cell transplant recipients, and patients with hematologic malignancies. Virulence. 2017. doi: 10.1080/21505594.2016.1213472 PubMed PMID: 27470662; PubMed Central PMCID: null. Barman P, Choudhary D, Chopra S, Thukral T. Blood stream infections in hematopoietic stem cell transplant patients: A 2-year study from India. Oncol J India. 2020;4(2):43. doi: 10.4103/oji.oji_7_20 Stecher B, Denzler R, Maier L, Bernet F, Sanders MJ, Pickard DJ, et al. Gut inflammation can boost horizontal gene transfer between pathogenic and commensal Enterobacteriaceae. Proc Natl Acad Sci. 2012;109(4):1269–74. doi: 10.1073/pnas.1113246109 Anthony WE, Burnham CAD, Dantas G, Kwon JH. The Gut Microbiome as a Reservoir for Antimicrobial Resistance. J Infect Dis. 2021;223(Supplement_3):S209–13. doi: 10.1093/infdis/jiaa497 Sardzikova S, Andrijkova K, Svec P, Beke G, Klucar L, Minarik G, et al. Gut diversity and the resistome as biomarkers of febrile neutropenia outcome in paediatric oncology patients undergoing hematopoietic stem cell transplantation. Sci Rep. 2024;14(1):5504. doi: 10.1038/s41598-024-56242-8 Iskandar K, Molinier L, Hallit S, Sartelli M, Hardcastle TC, Haque M, et al. Surveillance of antimicrobial resistance in low- and middle-income countries: a scattered picture. Antimicrob Resist Infect Control. 2021;10(1):63. doi: 10.1186/s13756-021-00931-w Ghazal SS, Stevens MP, Bearman GM, Edmond MB. Utility of surveillance blood cultures in patients undergoing hematopoietic stem cell transplantation. Antimicrob Resist Infect Control. 2014;3:20. doi: 10.1186/2047-2994-3-20 PubMed PMID: 24999384; PubMed Central PMCID: PMC4082284. Murray CJL, Ikuta KS, Sharara F, Swetschinski L, Robles Aguilar G, Gray A, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. The Lancet. 2022;399(10325):629–55. doi: 10.1016/S0140-6736(21)02724-0 Krakau S, Straub D, Gourlé H, Gabernet G, Nahnsen S. nf-core/mag: a best-practice pipeline for metagenome hybrid assembly and binning. NAR Genomics Bioinforma. 2022;4(1):lqac007. doi: 10.1093/nargab/lqac007 Gurevich A, Saveliev V, Vyahhi N, Tesler G. QUAST: quality assessment tool for genome assemblies. Bioinformatics. 2013;29(8):1072–5. doi: 10.1093/bioinformatics/btt086 Uritskiy GV, DiRuggiero J, Taylor J. MetaWRAP—a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome. 2018;6(1):158. doi: 10.1186/s40168-018-0541-1 Kang DD, Li F, Kirton E, Thomas A, Egan R, An H, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ. 2019;7:e7359. doi: 10.7717/peerj.7359 Wu YW, Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics. 2016;32(4):605–7. doi: 10.1093/bioinformatics/btv638 Alneberg J, Bjarnason BS, De Bruijn I, Schirmer M, Quick J, Ijaz UZ, et al. Binning metagenomic contigs by coverage and composition. Nat Methods. 2014;11(11):1144–6. doi: 10.1038/nmeth.3103 Chaumeil PA, Mussig AJ, Hugenholtz P, Parks DH. GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database. Hancock J, editor. Bioinformatics. 2020;36(6):1925–7. doi: 10.1093/bioinformatics/btz848 Wood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome Biol. 2019;20(1):257. doi: 10.1186/s13059-019-1891-0 Lu J, Breitwieser FP, Thielen P, Salzberg SL. Bracken: estimating species abundance in metagenomics data. PeerJ Comput Sci. 2017;3:e104. doi: 10.7717/peerj-cs.104 Lu J, Rincon N, Wood DE, Breitwieser FP, Pockrandt C, Langmead B, et al. Metagenome analysis using the Kraken software suite. Nat Protoc. 2022;17(12):2815–39. doi: 10.1038/s41596-022-00738-y Foster ZSL, Sharpton TJ, Grünwald NJ. Metacoder: An R package for visualization and manipulation of community taxonomic diversity data. Poisot T, editor. PLOS Comput Biol. 2017;13(2):e1005404. doi: 10.1371/journal.pcbi.1005404 Oksanen J, Simpson GL, Blanchet FG, Kindt R, Legendre P, Minchin PR, et al. vegan: Community Ecology Package [Internet]. 2001 [cited 2025 Nov 21]. p. 2.7-2. Available from: https://CRAN.R-project.org/package=vegan doi: 10.32614/CRAN.package.vegan Arbizu PM. pairwiseAdonis: pairwise multilevel comparison using adonis. 2017. 2019. Olm MR, Crits-Christoph A, Bouma-Gregson K, Firek BA, Morowitz MJ, Banfield JF. inStrain profiles population microdiversity from metagenomic data and sensitively detects shared microbial strains. Nat Biotechnol. 2021;39(6):727–36. doi: 10.1038/s41587-020-00797-0 Csardi MG. Package ‘igraph’. Last Accessed. 2013;3(09):2013. Pedersen TL. tidygraph: A Tidy API for Graph Manipulation [Internet]. 2023. Available from: https://github.com/thomasp85/tidygraph Si B, Liang Y, Zhao J, Zhang Y, Liao X, Jin H, et al. GGraph: An Efficient Structure-Aware Approach for Iterative Graph Processing. IEEE Trans Big Data. 2022;8(5):1182–94. doi: 10.1109/TBDATA.2020.3019641 Korula A, Perumalla S, Devasia AJ, Abubacker FN, Lakshmi KM, Abraham A, et al. Drug-resistant organisms are common in fecal surveillance cultures, predict bacteremia and correlate with poorer outcomes in patients undergoing allogeneic stem cell transplants. Transpl Infect Dis. 2020;22(3). doi: 10.1111/tid.13273 Wong SP, Er YX, Tan SM, Lee SC, Rajasuriar R, Lim YAL. Oral and Gut Microbiota Dysbiosis is Associated with Mucositis Severity in Autologous Hematopoietic Stem Cell Transplantation: Evidence from an Asian Population. Transplant Cell Ther. 2023;29(10):633.e1-633.e13. doi: 10.1016/j.jtct.2023.06.016 D’Amico F, Soverini M, Zama D, Consolandi C, Severgnini M, Prete A, et al. Gut resistome plasticity in pediatric patients undergoing hematopoietic stem cell transplantation. Sci Rep. 2019;9(1):5649. doi: 10.1038/s41598-019-42222-w Nieman AE, Savelkoul PHM, Beishuizen A, Henrich B, Lamik B, MacKenzie CR, et al. A prospective multicenter evaluation of direct molecular detection of blood stream infection from a clinical perspective. BMC Infect Dis. 2016;16(1):314. doi: 10.1186/s12879-016-1646-4 Balletto E, Mikulska M. Bacterial infections in haematopoietic stem cell transplant patients. Mediterr J Hematol Infect Dis. 2015;7:e2015045. doi: 10.4084/mjhid.2015.045 Jazmati T, Hamprecht A, Jazmati N. Comparison of stool samples and rectal swabs with and without pre-enrichment for the detection of third-generation cephalosporin-resistant Enterobacterales (3GCREB). Eur J Clin Microbiol Infect Dis. 2021;40(11):2431–6. doi: 10.1007/s10096-021-04250-1 Ghosh S, Bhattacharya S, Goel G, Deshmukh RA, Javed R, Roychowdhury M, et al. Hematopoietic stem-cell transplantation in a zoo of multidrug‐resistant organisms: Data from a cancer center in eastern India. Transpl Infect Dis. 2023;e14072. doi: 10.1111/tid.14072 Additional Declarations No competing interests reported. Supplementary Files HSCTsupplementarydatanpjantimicrobials.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9289820","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":628173124,"identity":"b5cfbb15-c047-4c6c-8397-a0771c38cfc2","order_by":0,"name":"Archana Madhav","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Archana","middleName":"","lastName":"Madhav","suffix":""},{"id":628173125,"identity":"34be463a-f922-471c-b281-62c81c433500","order_by":1,"name":"Jobin John Jacob","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jobin","middleName":"John","lastName":"Jacob","suffix":""},{"id":628173126,"identity":"d25c3e03-8240-4b0d-8e86-9ee36a2cb5e0","order_by":2,"name":"Sushil Selvarajan","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Sushil","middleName":"","lastName":"Selvarajan","suffix":""},{"id":628173127,"identity":"865778c7-7a09-401b-9de1-421dcf1e8a77","order_by":3,"name":"Sanika Kulkarni","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Sanika","middleName":"","lastName":"Kulkarni","suffix":""},{"id":628173128,"identity":"ddd1f718-e118-4b02-ba87-2d959c1bcb6a","order_by":4,"name":"Sourav Das","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Sourav","middleName":"","lastName":"Das","suffix":""},{"id":628173130,"identity":"33048e08-2545-410d-914a-548285310961","order_by":5,"name":"Dhivya Murugan","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Dhivya","middleName":"","lastName":"Murugan","suffix":""},{"id":628173132,"identity":"26bad5ae-2800-45fe-bb1f-2dc50c7e8746","order_by":6,"name":"Pavithra Sathiya Narayanan","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Pavithra","middleName":"Sathiya","lastName":"Narayanan","suffix":""},{"id":628173134,"identity":"09fc36b1-a4fd-4e5f-afa3-340e6f08ddb2","order_by":7,"name":"Yesudoss Muniyandi","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yesudoss","middleName":"","lastName":"Muniyandi","suffix":""},{"id":628173135,"identity":"41ce9076-e7a2-412a-8c7d-7e9e3c545689","order_by":8,"name":"Kalaiyarasi Mohan","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Kalaiyarasi","middleName":"","lastName":"Mohan","suffix":""},{"id":628173139,"identity":"904faf65-9aec-43d6-a858-23ccb9bf9a5f","order_by":9,"name":"Biju George","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Biju","middleName":"","lastName":"George","suffix":""},{"id":628173140,"identity":"08099980-988f-423d-9c3d-075773f3ae25","order_by":10,"name":"Balaji Veeraraghavan","email":"","orcid":"","institution":"Christian Medical College \u0026 Hospital","correspondingAuthor":false,"prefix":"","firstName":"Balaji","middleName":"","lastName":"Veeraraghavan","suffix":""},{"id":628173141,"identity":"fd618957-068f-4e52-967f-cdadb7e446b0","order_by":11,"name":"Stephen Baker","email":"data:image/png;base64,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","orcid":"","institution":"Agency for Science, Technology and Research","correspondingAuthor":true,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Baker","suffix":""}],"badges":[],"createdAt":"2026-04-01 09:38:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9289820/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9289820/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107838390,"identity":"7cd70c7f-b92e-4a26-a261-f53166422fb3","added_by":"auto","created_at":"2026-04-26 17:10:37","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1141662,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMetacoder plots showing the abundance of reads belonging to various taxa in the HSCT study. \u003c/strong\u003eFigure depicts an overview of the collective microbial composition of the MacConkey-enriched stool sample collection used in the study. The detected taxids are shown in the node colour with the distribution of reads belonging to each taxon shown in the branch colours. \u003cstrong\u003eEnlarged Pseudomonadota (Proteobacteria) (A) and Bacillota (Firmicutes) (B) phyla. \u003c/strong\u003ePlate-sweep selection for bacterial species within two main classes – Gammaproteobacteria (Pseudomonadota), and Bacilli (Bacillota), the latter mainly comprising Staphylococci of the Bacillales order and Enterococci of the\u003cstrong\u003e \u003c/strong\u003eLactobacillales order.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9289820/v1/13a2ebc2f654fdc8dff05148.png"},{"id":107869950,"identity":"bae710b5-4556-4be1-b8be-eeb4a44d428b","added_by":"auto","created_at":"2026-04-27 07:38:32","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":386541,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e(A) NMDS plot depicting beta-diversity of enriched stool samples using the Bray-Curtis dissimilarity index.\u003c/strong\u003e Plot shows indices coloured by the timepoint when the sample was collected, relative to when the patient’s transplant was performed. Plot shows clustering of samples collected during the conditioning week (orange), and a looser clustering of samples collected during the first two-weeks post stem-cell infusion (blue; pre-engraftment). (\u003cstrong\u003eB) Volcano plot depicting log2 fold-changes in enriched taxa at genus level between patients with and without GI-GvHD at the pre-engraftment timepoint.\u003c/strong\u003eTaxa in the top right quadrant were detected in significantly higher abundance among patients who went on to develop GI-GvHD, especially the \u003cem\u003eEnterococcus\u003c/em\u003eand \u003cem\u003eMorganella\u003c/em\u003e genera. While there were other genera (e.g., \u003cem\u003eLuteimonas\u003c/em\u003e, \u003cem\u003eCandidatus\u003c/em\u003e \u003cem\u003ehamiltonella\u003c/em\u003e) showing log-fold shifts, these were not considered to be representative shifts relative to the residual microbiome content, since we did not enrich for these organisms.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9289820/v1/e90ae2544f77c72740f1c058.jpeg"},{"id":107870463,"identity":"f02ee3a3-174b-4d85-b24a-ed71fb45035f","added_by":"auto","created_at":"2026-04-27 07:39:43","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":960596,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA) Boxplot showing the Shannon alpha-diversity indexes for all AMR genes detected among the sweep-enriched taxa longitudinally through the transplant timeline.\u003c/strong\u003e Overall AMR gene diversity saw a significant decrease (\u003cem\u003ep\u003c/em\u003e\u0026lt;0.001***) when comparing samples taken during the conditioning week to those taken in the pre-engraftment period, and a significant increase (\u003cem\u003ep\u003c/em\u003e\u0026lt;0.001***) when comparing pre-engraftment samples to early post-engraftment samples. \u003cstrong\u003eB) Boxplot showing the Shannon alpha-diversity indexes for non-efflux AMR genes detected among sweep-enriched taxa longitudinally through the transplant timeline.\u003c/strong\u003e Alpha diversity of non-efflux AMR genes showed no significant differences when comparing samples taken during the conditioning week to those taken in the pre-engraftment period, but significant increases were observed when comparing pre-engraftment samples to those collected at successive early post-engraftment (**\u003cem\u003ep\u003c/em\u003e\u0026lt;0.01) and late post-engraftment (*\u003cem\u003ep\u003c/em\u003e\u0026lt;0.05) timepoints.\u003cstrong\u003e C) Longitudinal variations in the abundance of 5 non-efflux MDR-conferring AMR genes of clinical relevance detected per enriched sample.\u003c/strong\u003e There was a significant increase in ESBL gene abundance at the early post-engraftment timepoint compared to the conditioning week and pre-engraftment timepoints. There were progressive increases in the number of \u003cem\u003eblaNDM\u003c/em\u003eand \u003cem\u003eblaOXA\u003c/em\u003e genes detected per sample when comparing conditioning week samples to pre-engraftment and early post-engraftment samples. \u003cstrong\u003eD) Hi-C metagenomic sequencing was performed on a subset of 11 enriched stool samples from patients with clinically significant infection episodes.\u003c/strong\u003eHi-C helped delineate AMR genes that co-localised on plasmids. The heatmap shows a presence/absence matrix of AMR genes found in genomic regions and on plasmids from each of the 11 enriched stool samples which underwent Hi-C.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9289820/v1/f5a08b1dbd9df4cb83745b42.jpeg"},{"id":107838394,"identity":"5d1b2b02-1b8d-466f-b0c8-3c1e75d8de1c","added_by":"auto","created_at":"2026-04-26 17:10:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":187101,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMicrobial network depicting highly similar isolates shared between patients or strains detected within participants across multiple timepoints.\u003c/strong\u003e Figure shows plate-sweep enriched stool reads which mapped to reconstructed MAG reference bins of targeted taxa with at least \u0026gt;70% completeness and \u0026lt;5% contamination, at 99.999% ANI and where \u0026gt;25% of the genome was mapped.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9289820/v1/9dc871d34f86ac706b1002e4.png"},{"id":109019010,"identity":"b9fd417e-6d05-4f9f-b8e6-bcf77f3f8910","added_by":"auto","created_at":"2026-05-11 18:26:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2743477,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9289820/v1/c9b2c671-5e25-49ea-b34a-3083eeffd72c.pdf"},{"id":107870761,"identity":"9b9c9b37-26aa-4425-b186-bad9cde25bcf","added_by":"auto","created_at":"2026-04-27 07:40:34","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2351706,"visible":true,"origin":"","legend":"","description":"","filename":"HSCTsupplementarydatanpjantimicrobials.docx","url":"https://assets-eu.researchsquare.com/files/rs-9289820/v1/09fa0bf097dca625b193ec73.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Stool-based genomic surveillance identifies post-engraftment expansion of multidrug-resistant pathogens in haematopoietic stem cell transplant patients in India","fulltext":[{"header":"Introduction","content":"\u003cp\u003eBetween 15\u0026ndash;65% of all haematopoietic stem cell transplantation (HSCT) patients develop bacterial bloodstream infections (BSI), two-thirds of which are associated with mucosal barrier injury (MBI) (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). \u003cem\u003eEnterobacteriaceae\u003c/em\u003e are the most common aetiologic agents of Gram-negative bacteraemia in HSCT patients (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), accounting for ~\u0026thinsp;75% of culture-positive BSIs post-HSCT in India (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). Risk of \u003cem\u003eEnterobacteriaceae\u003c/em\u003e BSIs is amplified by their potential to accumulate antimicrobial resistance (AMR) determinants, which may be exacerbated by stress induced by intestinal inflammation (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e), with the gut acting as a reservoir of AMR genes (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). Sardzikova \u003cem\u003eet al.\u003c/em\u003e reported that multi-drug resistant (MDR) \u003cem\u003eE. coli\u003c/em\u003e, \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e, \u003cem\u003eEnterococcus faecium\u003c/em\u003e and \u003cem\u003eStaphylococcus aureus\u003c/em\u003e are leading contributors to the resistome in the HSCT gut (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eLow-middle income countries (LMICs) consistently report a higher prevalence of AMR in almost all bacteria, yet syndrome-specific data (e.g., AMR in patients with BSIs post-HSCT) are scarce (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). The cocktail of immunosuppressants and antimicrobials administered post-HSCT reduce the sensitivity of culture-based diagnostics of BSI, thus limiting the application of antimicrobial susceptibility testing (AST) (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). The growing challenge posed by AMR in LMICs (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e), coupled with the limited HSCT patient data in this region, underscores the need for further investigation, where genomic approaches could offer new insights and solutions.\u003c/p\u003e \u003cp\u003eHere, we enrolled HSCT patients at a tertiary care centre in India and utilised plate-sweep enrichment on longitudinally collected stool samples. We targeted priority AMR pathogens frequently associated with BSIs post-HSCT and their associated resistome trajectories. We additionally leveraged enhanced metagenome-assembled genome (MAG) recovery from enriched samples and conducted whole-genome sequencing (WGS) of bloodstream isolates from BSI episodes to detect organisms co-present in the gut and bloodstream of patients with culture-positive BSI. Lastly, we derived plasmid-host associations for patients with culture-positive BSIs post-HSCT using Hi-C on the sweep-enriched fraction of stool samples and tracked sweep-enriched strains within and between participants in our cohort.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy design\u003c/h2\u003e \u003cp\u003e We conducted a prospective longitudinal study enrolling 81 participants undergoing allogeneic HSCT between May and December 2022 at the Department of Haematology, Christian Medical College (CMC), Vellore, India. The study was a collaboration between the Cambridge Institute for Therapeutic Immunology and Infectious Disease (CITIID) and the Clinical Microbiology and Clinical Haematology departments at CMC Vellore. This study was granted ethical approval by the Institutional Review Board (IRB), Research and Ethics committee at CMC Vellore in accordance with the Declaration of Helsinki and the Indian Council of Medical Research National Ethical Guidelines for Biomedical and Health Research involving Human Participants (IRB Min No. 14605 dated 27.04.2022).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eParticipant eligibility criteria and clinical data management\u003c/h3\u003e\n\u003cp\u003ePatients admitted to the Department of Haematology at CMC Hospital were eligible if scheduled to undergo an allogeneic stem cell transplant and provided informed consent to participate. Informed consent forms were written in English and translated into Hindi and Tamil. A clinical case report form recorded clinical metadata including age, sex, diagnosis, donor age, donor sex, HLA match, type of transplant, preparative regimen, post-HSCT infection episodes, episodes of \u003cem\u003eClostridioides difficile\u003c/em\u003e colitis, graft-versus-host disease (GvHD) and mucositis, date of absolute neutrophil count (ANC) recovery, total days of hospital stay, and outcome. Additional clinical information was obtained from the clinicians or nurses in-charge, accessed through the clinical workstation, or through discharge/death summaries. Sensitive clinical information was accessed through secure hospital servers and shared only with key personnel.\u003c/p\u003e\n\u003ch3\u003eStool sample collection and processing\u003c/h3\u003e\n\u003cp\u003eStool samples were collected in sterile stool collection containers and transferred to the Department of Clinical Microbiology at 4\u0026ordm;C. Once received, samples were labelled and processed within 72 hours. Blood samples were collected for culture and AST if a BSI was suspected by the clinician in-charge, for example, if the patient developed a fever post-HSCT.\u003c/p\u003e \u003cp\u003eStool samples were homogenised in a 1:1 ratio with PBS and 100x dilutions were plated on MacConkey agar plates without crystal violet to grow Gram-negative bacteria and Gram-positive bacteria belonging to the \u003cem\u003eStaphylococcus\u003c/em\u003e and \u003cem\u003eEnterococcus\u003c/em\u003e genera. Following overnight incubation, a complete sweep of the plate was collected by resuspending any grown colonies in sterile PBS. DNA extraction was done on all plate sweep aspirates using the Wizard\u0026reg; Genomic DNA purification kit by Promega per the manufacturer\u0026rsquo;s protocol. A subset of plate sweep aspirates (selected retrospectively for patients with post-HSCT culture positive BSI) additionally underwent formalin inactivation for Hi-C. Detailed protocols for stool sample processing, serial dilutions, plate sweeps, DNA extraction and formalin inactivation can be found in supplementary methods.\u003c/p\u003e\n\u003ch3\u003eBlood sample collection and processing\u003c/h3\u003e\n\u003cp\u003eBlood culture bottles were incubated in the BacT/ALERT\u0026reg; automated Blood Culture system. Positive samples were passaged either on blood agar or selective media chosen according to the pathogen detected to obtain pure isolates for WGS. Isolates were incubated overnight in nutrient broth at 37\u0026ordm;C, and a 1 mL aliquot of the overnight culture was used for DNA extraction and WGS.\u003c/p\u003e\n\u003ch3\u003eGenome sequencing\u003c/h3\u003e\n\u003cp\u003eDNA concentrations were quantified on a Qubit\u0026trade; 2.0 Fluorometer, and shipped to Eurofins Genomics for standard library preparation with equimolar pooling and sequencing. Samples were sequenced on an Illumina NovaSeq 6000 S4 PE150 instrument, with a minimum of 10\u0026nbsp;million paired-end reads generated for whole metagenomic sequencing (WMS) of enriched stool samples and 100x coverage for whole genome sequencing (WGS) for bloodstream isolates. The sequencing reads for both stool and blood samples from the study can be found under BioProject accession PRJNA1072756.\u003c/p\u003e \u003cp\u003eTwo-hundred and ninety-eight (298) enriched stool samples were sent for WMS. Twenty samples contained insufficient DNA to generate QC-checked libraries (\u0026lt;\u0026thinsp;1nM/L of DNA). After removing samples with incomplete metadata or belonging to patients excluded based on patient eligibility criteria, a total of 252 samples from 81 patients were included in the study.\u003c/p\u003e \u003cp\u003eFor Hi-C, 13 formalin-inactivated enriched stool plate sweeps were sent to Phase Genomics in Seattle, USA for Hi-C library prep and sequencing according to the ProxiMeta\u003csup\u003e\u0026trade;\u003c/sup\u003e Hi-C Kit Protocol v4.5. Contigs assembled from enriched stool metagenomic read data were sent to the bioinformatics team at Phase Genomics to overlay Hi-C connections. Two stool samples (P14AT0065 and P51AT1247) did not pass QC for Hi-C.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eQuality control and assembly of sequencing reads\u003c/h2\u003e \u003cp\u003eRaw sequencing reads from the enriched stool samples were run through the nf-core/mag pipeline v2.3.0 (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e) for trimming, adapter removal, and quality control with FastQC. Enriched stool metagenomic read assembly was performed using SPAdes v3.15.4, and all assemblies underwent QC using QUAST (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). We used Metawrap v1.3.0 (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e) to produce consensus bins from three binning pipelines, MetaBAT2 v2.12.1 (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e), Maxbin2 v2.2.7 (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e) and CONCOCT v1.1.0 (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). The threshold for consensus bins was set to a minimum of 70% completeness and \u0026lt;\u0026thinsp;5% contamination to extract bins of high purity. After extracting consensus bins, each bin was passed through GTDB-TK (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e) for species identification.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eTaxonomic assignment and microbial diversity\u003c/h3\u003e\n\u003cp\u003eTaxonomic assignment was performed directly on enriched stool plate sweep reads using Kraken v2.1.2 (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e), followed by Bayesian re-estimation at genus level using Bracken v2.7.0 (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e). Residual human OTUs were removed with the \u0026ldquo;extract_kraken_reads.py\u0026rdquo; script. Bracken outputs were combined using the \u0026ldquo;combine_kraken_reports.py\u0026rdquo; script and converted into a .biom format using kraken-biom v1.2.0 from the KrakenTools software suite (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). OTU tables were imported into RStudio v4.2.3 and visualisations for taxonomic composition of the enriched fraction of stool samples was performed using metacoder v0.3.6 (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). Beta diversity was calculated using the Bray-Curtis dissimilarity index with the vegdist function from the vegan package v2.5-7 (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e). A pairwise adonis test was performed using the pairwiseAdonis v0.4.1 (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e) package on R.\u003c/p\u003e\n\u003ch3\u003eAMR gene detection and antimicrobial sensitivity testing\u003c/h3\u003e\n\u003cp\u003eAMR genes were detected from bloodstream isolates and enriched stool sequencing data using ABRIcate v1.0.1 with the CARD database as reference with default parameters. Antimicrobial drug classes were reclassified as \u0026lsquo;multidrug\u0026rsquo; if the AMR gene detected conferred resistance to \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\ge\\:\\)\u003c/span\u003e\u003c/span\u003e3 antimicrobial classes.\u003c/p\u003e \u003cp\u003ePhenotypic AST was performed on bloodstream isolates and stool colony picks using 6mm disks of Whatmann No. 1 filter paper saturated with antimicrobial solutions. Stool colony picks were tested against 8 antimicrobials for Gram-negative organisms and 4 antimicrobials for Gram-positive organisms. A complete description of drug-brug AST combinations for stool colony picks is provided in supplementary methods.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eInStrain analysis and microbial network\u003c/h2\u003e \u003cp\u003eInStrain v1.8.0 (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e) was used in this study for two objectives 1) to identify episodes where bloodstream isolates from patients with culture positive BSI post-HSCT were also present in the gut, suggesting translocation, and 2) to detect strains from priority AMR organisms shared across multiple enriched stool samples. Correspondingly, we used two different ANI thresholds when attempting to detect isolates across samples. For detecting possible translocation events, we used a 99.9% population-level ANI to account for variation introduced by differences in sample processing and sequencing between blood and stool samples and a 50% benchmarked breadth threshold as recommended in the inStrain pipeline documentation to minimise risk of false positives. For detecting shared isolates across stool samples, we used thresholds of 99.999% population-level ANI and at least 25% \u0026lsquo;genome compared\u0026rsquo; between samples. Details of inStrain pipeline configuration and output analyses can be found in supplementary methods.\u003c/p\u003e \u003cp\u003eThe microbial network depicting strains from targeted AMR organisms shared across multiple enriched stool samples was created using the igraph v1.5.1 (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e), tidygraph v1.2.3 (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e) and ggraph v2.1.0 (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e) packages on RStudio. Each node (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;98) represents an enriched stool sample and edges between nodes denote connections (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;174), i.e., bacterial strains shared between samples. The number of connections to each sample was used to produce a network plot using ggraph under the \u0026lsquo;kk\u0026rsquo; layout.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCohort summary and outcomes\u003c/h2\u003e \u003cp\u003eA summary of the baseline clinical characteristics of the 81 patients in the allogeneic HSCT cohort is shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Overall, 32/81 patients developed a clinically significant bacterial infection (38 episodes total) post-HSCT. \u003cem\u003eClostridioides difficile\u003c/em\u003e colitis was the most common infection post-HSCT (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;11), followed by \u003cem\u003eEnterobacteriaceae\u003c/em\u003e (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7) and coagulase negative \u003cem\u003eStaphylococcus\u003c/em\u003e (CONS) (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7, including 2 episodes of central-line associated BSI). Excluding catheter and central-line associated BSI (n\u0026thinsp;=\u0026thinsp;3), 17 patients experienced culture-positive bacterial BSIs post-HSCT (20 episodes in total). There were 10 episodes of culture-positive BSI prior to HSCT among seven patients (resolved before transplantation).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline characteristics of the HSCT study cohort\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eBaseline characteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cem\u003en\u003c/em\u003e (%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eAge (years)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e38 (46.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18\u0026ndash;30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (16.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31\u0026ndash;40\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (16.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41\u0026ndash;50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11 (13.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51\u0026ndash;60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6 (7.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSex (% male)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e53 (65.4%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDiagnosis\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMalignant\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e39 (48.1%\u003c/b\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAcute lymphoblastic leukaemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (33.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAcute myeloid leukaemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9 (23.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMyelodysplastic syndrome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8 (20.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eChronic myeloid leukaemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5 (12.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHodgkin's lymphoma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2 (5.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePolycythemia vera\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNK T-cell lymphoma\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eNon-malignant\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e42 (51.9%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThalassemia major\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17 (40.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSevere aplastic anaemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15 (35.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFanconi anaemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4 (9.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSickle cell anaemia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWiskott Aldrich syndrome\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMucopolysaccharidosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eChronic granulomatous disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSevere congenital neutropenia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eParoxysmal nocturnal haemoglobinuria\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (2.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eType of allogeneic transplant\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eFully matched\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e51 (63.0%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRelated donor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e40 (49.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUnrelated donor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11 (13.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHaploidentical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e30 (37.0%)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eConditioning regimen\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMyeloablative\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58 (71.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eReduced intensity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23 (28.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOutcome\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDischarged\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66 (81.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDied\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15 (18.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe median time to culture-positive BSI was 12 days post-HSCT. We recovered 19 bacterial BSI isolates on culture plates, which were subjected to WGS (7/10 pre- and 12/20 post-HSCT). Gram-negative bacilli (GNB) caused nearly two-thirds of post-transplant BSI episodes (13/20). The mortality rate among patients who developed culture-positive BSI post-HSCT was 41.2% (7/17), whereas the mortality rate among patients who developed any infectious complication post-HSCT was 37.5% (12/32). Supplementary data (SD) 1 contains a summary of all clinically significant organisms detected post-HSCT in our cohort.\u003c/p\u003e \u003cp\u003eFifty-seven patients had mucositis (median grade 3). Out of 81 patients, 21 experienced GvHD, of whom 15 had gastrointestinal (GI)-GvHD. Over half of all patients who had GI-GvHD (8/15) also had episodes of \u003cem\u003eClostridioides difficile\u003c/em\u003e infection. Patients 12, 18 and 26 had haploidentical HSCTs, developed overall GvHD at grade 4, and all three had \u003cem\u003eClostridioides difficile\u003c/em\u003e infections. Patients 12 and 18 succumbed to complications they developed 51- and 77-days post-admission respectively, but patient 26 recovered and was discharged 93 days after admission.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003ePlate-sweep enrichment enables high-resolution recovery of priority AMR pathogens\u003c/h2\u003e \u003cp\u003eA median of three stool samples was collected from each patient (range 1\u0026ndash;8), with at least three stool samples collected for 61/81 patients. Stool samples (n\u0026thinsp;=\u0026thinsp;252) were categorised into five groups based on the timepoint relative to HSCT: pre-conditioning (before day\u0026thinsp;\u0026minus;\u0026thinsp;8; \u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;12), conditioning week (day\u0026thinsp;\u0026minus;\u0026thinsp;8 to day 0; \u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;78), pre-engraftment (day\u0026thinsp;+\u0026thinsp;1 to day\u0026thinsp;+\u0026thinsp;15; \u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;89), early post-engraftment (day\u0026thinsp;+\u0026thinsp;16 to day\u0026thinsp;+\u0026thinsp;100; \u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;67) and late post-engraftment (day\u0026thinsp;+\u0026thinsp;101 to day\u0026thinsp;+\u0026thinsp;200; \u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;6).\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e outlines the bacterial taxa that were enriched in the MacConkey plate sweeps. We specifically targeted Gram-negative organisms (i.e., \u003cem\u003eE. coli\u003c/em\u003e, \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e, \u003cem\u003eAcinetobacter baumannii\u003c/em\u003e and \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e) but the omission of crystal violet permitted the growth of \u003cem\u003eStaphylococcus\u003c/em\u003e spp. and \u003cem\u003eEnterococcus\u003c/em\u003e spp. Therefore, apart from \u003cem\u003eClostridioides difficile\u003c/em\u003e, our MacConkey plate sweep enrichment approach supported the detection of all clinically significant organisms that emerged in our HSCT cohort (SD1). We intentionally selected for priority AMR pathogens in the HSCT cohort with this approach, as described by the focused colour dispersion within specific branches of Pseudomonadota (Proteobacteria; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) and Bacillota (Firmicutes; Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe 252 stool samples generated between 1.8 to 33\u0026nbsp;million reads per sample, with nearly 3.5\u0026nbsp;billion individual operational taxonomic units (OTUs) spanning 4,272 taxonomic levels. This dataset was rarefied at the 10th percentile, representing approximately half of the original dataset with over 1.6\u0026nbsp;billion OTUs. Over three-quarters (77.6%; \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1,282,928,757) of the OTUs belonged to the \u003cem\u003eEnterobacteriaceae\u003c/em\u003e family in the Pseudomonadota (Proteobacteria) phylum. At genus level, this family mainly consisted of \u003cem\u003eEscherichia\u003c/em\u003e (73.4%), \u003cem\u003eKlebsiella\u003c/em\u003e (21.8%), \u003cem\u003eShigella\u003c/em\u003e (1.54%), \u003cem\u003eCitrobacter\u003c/em\u003e (1.38%) and \u003cem\u003eEnterobacter\u003c/em\u003e (1.34%). In the Bacillota (Firmicutes) phylum (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;217,106,676), Lactobacillales and Bacillales orders comprised 89.6% and 10.4% of OTUs in the phylum, respectively. The Lactobacillales order was dominated by \u003cem\u003eEnterococcus\u003c/em\u003e spp. (99.8%), and the genus \u003cem\u003eStaphylococcus\u003c/em\u003e was the most abundant (99.6%) in the Bacillales order. The metacoder plot hence depicts a selectively enriched subcommunity of organisms from stool samples that have previously been identified as leading contributors to AMR and BSI among HSCT patients and notably does not capture total gut microbiome composition.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eTaxonomic diversity of enriched stool from HSCT patients\u003c/h2\u003e \u003cp\u003eA Shannon alpha diversity analysis of enriched stool samples at species-level, as expected, did not show any significant difference (one-way ANOVA) when grouped by timepoints relative to HSCT (SD2), suggesting even enrichment of the intended subcommunity across samples. Beta diversity was calculated to examine the population level differences in the enriched subcommunities relative to the transplant timeline using the Bray-Curtis dissimilarity index (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). The plot shows clustering of conditioning week samples (orange). A pairwise adonis test was used to probe grouped timepoint beta diversities in a pairwise manner (***\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; SD3). These data depicted a notable restructure among the enriched subcommunity during the conditioning week when compared to samples from the pre-engraftment (R\u0026sup2; 0.156), early- (R\u0026sup2; 0.105), and late post-engraftment timepoints (R\u0026sup2; 0.161); the genera contributing to these differences are shown in SD4. No significant associations were observed when assessing beta diversity of the enriched subcommunity in relation to transplant outcomes or GvHD. However, patients who developed acute GvHD post-HSCT demonstrated significant log\u003csub\u003e2\u003c/sub\u003e fold-increases in the \u003cem\u003eEnterococcus\u003c/em\u003e (***\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and \u003cem\u003eMorganella\u003c/em\u003e (***\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) genera (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eTemporal dynamics of AMR in target organisms\u003c/h2\u003e \u003cp\u003eABRIcate was used to detect AMR genes among the enriched taxa from the HSCT cohort\u0026rsquo;s stool samples using the CARD database. 16,903 AMR gene hits were detected among the 252 enriched stool samples, with efflux-mediating genes representing 65.8% (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;11,118) of AMR genes detected within the collection. The median number of AMR genes detected among enriched taxa in the pre-conditioning phase (\u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;12) and conditioning week (\u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;78) were 62 and 68 AMR genes per sample, respectively, compared to a median of 55 AMR genes detected per enriched sample during pre-engraftment (\u003cem\u003en\u0026thinsp;=\u003c/em\u003e\u0026thinsp;89). Correspondingly, a median Shannon alpha diversity index of 3.23 was observed for AMR genes during both the pre-conditioning and conditioning phases (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). At the pre-engraftment timepoint, the Shannon index dropped to 3.06. These decreases in the abundance and diversity of AMR genes between the conditioning week and pre-engraftment timepoints were both significant at ***\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001 with a pairwise t-test with Bonferroni correction (SD5).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAt early post-engraftment, there was a significant increase in the abundance (**\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and diversity (***\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) of AMR genes compared to the pre-engraftment timepoint among the enriched taxa. The median number of AMR genes detected per enriched sample at early post-engraftment was 69 (n\u0026thinsp;=\u0026thinsp;67; alpha diversity\u0026thinsp;=\u0026thinsp;3.32) and 96 at late post-engraftment (n\u0026thinsp;=\u0026thinsp;6; alpha diversity\u0026thinsp;=\u0026thinsp;3.46). The abundance of non-efflux AMR genes within the enriched subcommunity was significantly higher at early (**\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and late (*\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) post-engraftment compared to pre-engraftment, suggesting that organisms within the enriched subcommunity carrying non-efflux AMR genes remained in the HSCT gut post-transplantation. This observation suggests preferential selection through robust tolerance and survival mechanisms among these organisms but should be interpreted with caution since only six samples were collected at late post-engraftment.\u003c/p\u003e \u003cp\u003eThe abundance of five MDR-conferring ESBL/carbapenemase genes (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;695) \u0026ndash; \u003cem\u003eblaOXA\u003c/em\u003e, \u003cem\u003eblaNDM\u003c/em\u003e, \u003cem\u003eblaCTX-M\u003c/em\u003e, \u003cem\u003eblaSHV\u003c/em\u003e and \u003cem\u003eblaTEM\u003c/em\u003e were plotted longitudinally against the HSCT timeline in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC. At least 1/5 MDR genes were detected in 216/252 enriched stool samples. Overall, we observed a significant successive increase in the abundance of MDR genes when comparing enriched samples collected during the early post engraftment timepoint to MDR genes detected in enriched samples during the preceding conditioning week (***\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and the pre-engraftment (**\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) timepoints. Notably, the \u003cem\u003eblaOXA\u003c/em\u003e and \u003cem\u003eblaNDM\u003c/em\u003e carbapenemase genes exhibited progressive increases in their mean abundance per sample between the conditioning week (0.60 and 0.23, respectively) and late post-engraftment (1.40 and 0.8, respectively). In multivariable logistic regression, most AMR genes were not associated with infection, except for a ~5x increased odds of infection among patients who carried the \u003cem\u003eblaNDM\u003c/em\u003e gene during the conditioning week (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042; SD6).\u003c/p\u003e \u003cp\u003eBy using a Hi-C approach on the sweep-enriched fractions of stool samples obtained from patients with culture positive BSI post-HSCT, we found that \u003cem\u003eblaOXA\u003c/em\u003e and \u003cem\u003eblaNDM\u003c/em\u003e genes often co-localised on plasmids harboured by \u003cem\u003eE. coli\u003c/em\u003e and \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e. Such plasmids were detected in half of all patients who developed culture positive BSIs; however, with non-identical AMR gene compositions (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). These plasmids also carried genes inducing resistance against tetracyclines, trimethoprim, sulphonamides, and macrolides. Among the 11 samples which underwent Hi-C, 70.3% of AMR genes detected were of plasmid origin; the plasmid-associated AMR gene to host summary is provided in SD7. AST data for select enteric MDR bacteria (\u003cem\u003eE. coli\u003c/em\u003e, \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e and \u003cem\u003eEnterococcus faecium\u003c/em\u003e) isolated from stool colony picks (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;90) alongside a subset of AMR genes from enriched stool samples are depicted in SD8. MDR isolates were detected in 57 samples from 41 patients, signifying that \u0026gt;\u0026thinsp;50% of patients had at least one phenotypically MDR organism isolated from their stool during the HSCT timeline.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eSpatiotemporal isolate tracking reveals persistence and possible transmission\u003c/h2\u003e \u003cp\u003ePlate-sweep metagenomics offers a factor-increase in the resolution of MAGs recovered from stool samples. This approach enabled us to probe microdiversity within enriched taxa to identify strains shared across multiple patients within the cohort, and map bloodstream isolates from clinical infection episodes which underwent WGS to enriched stool samples. Figure\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows a microbial network which depicts (at strain level) organisms from the selective subcommunity that were detected across multiple stool samples.\u003c/p\u003e \u003cp\u003eTo detect the same isolate among multiple participants, or across multiple timepoints within the same participant, we used a stringent average nucleotide identity (ANI) threshold of 99.999% at the population-level (popANI), where at least 25% of the enriched stool reads were compared against the reference consensus MAG bins extracted from the plate sweep sequencing data. The resulting microbial network consisted of 98 nodes (samples) and 174 edges (strains); the degree annotation represents the number of edges (i.e., strains) that connect to each node (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The network depicts clusters where sequencing reads from the enriched stool samples mapped with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\ge\\:\\)\u003c/span\u003e\u003c/span\u003e99.999% popANI to reconstructed MAG reference bins, indicating highly similar isolates detected across multiple samples. There were several episodes (participants 3, 11, 46, 54, 57, and 61) where isolates could be linked longitudinally within the same participant, and one participant (participant 36) who had MAGs from \u003cem\u003eEscherichia coli\u003c/em\u003e and \u003cem\u003eProvidencia alcalifaciens\u003c/em\u003e (\u0026gt;\u0026thinsp;85% complete, \u0026lt;5% contaminated) that could be linked across timepoints. Transmission links where two edges were shared between samples indicate episodes where reads from both enriched stool samples mapped to reconstructed MAGs from both samples at the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\ge\\:\\)\u003c/span\u003e\u003c/span\u003e99.999% ANI and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\ge\\:\\)\u003c/span\u003e\u003c/span\u003e25% \u0026lsquo;genome compared\u0026rsquo; thresholds. Cases where these bidirectional relationships appeared across multiple participants likely indicate localised transmission clusters but would need confirmation with environmental surveillance data.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eGenomic evidence links gut colonisation to bloodstream infections in select cases\u003c/h2\u003e \u003cp\u003eThe cultured organism from the blood samples was matched to an organism present in the selectively enriched stool fraction in five patients at strain-level \u0026ndash; patient 2 (\u003cem\u003eE. coli\u003c/em\u003e), 5 (\u003cem\u003eEnterococcus faecium\u003c/em\u003e), 14 (\u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e), 44 (\u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e) and 55 (\u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e). All five patients presented with malignant indications for HSCT, underwent myeloablative conditioning, and developed mucositis (median grade 3). Patients 2, 5, 14 and 44 had episodes of culture-positive BSI post-HSCT, whereas patient 55 had culture-positive BSI 45 days pre-HSCT. In patients 2, 5, and 14, the blood culture isolate was matched to enriched stool collected prior to BSI, suggesting temporally plausible gut-blood translocation, and in patients 44 and 55, the BSI isolate was matched to enriched stool collected 16 and 100 days after BSI, i.e., temporally ambiguous. In patient 55, the blood culture and AST report identified the organism as carbapenem-resistant \u003cem\u003eKlebsiella pneumoniae.\u003c/em\u003e After HSCT, patient 55 developed febrile neutropenia with septic shock and grade 3 mucositis, but empirical treatment with meropenem/colistin/teicoplanin likely contributed to subsequent sterile blood cultures. While the clinical infection in patient 55 occurred 45 days before HSCT and was resolved prior to transplantation, the post-HSCT detection of the BSI-causing \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e strain in the enriched stool sweep suggests that the organism persisted within the host in the gut. In 15.8% (3/19) of sequenced BSI cases, we obtained direct genomic evidence of prior gut colonisation by the causative pathogen. All five BSI isolates that were co-detected in the gut were genotypically MDR and four were phenotypically MDR (SD9); patient 14 died before AST could be performed. A clinical summary of the patients with BSI isolates mapping to a corresponding enriched stool bin is provided in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eSummary of bloodstream isolates from BSI episodes which mapped to the corresponding participant\u0026rsquo;s gut microbiome sample.\u003c/b\u003e Bloodstream isolates underwent WGS where reads mapped to reconstructed MAGs (\u0026gt;\u0026thinsp;97% completeness, \u0026lt;\u0026thinsp;2% contamination) from the enriched gut microbiome samples of the same participant at \u0026gt;\u0026thinsp;99.9% ANI across \u0026gt;\u0026thinsp;50% breadth of the reconstructed MAG.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ebloodstream_isolate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003egenome\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ebin_completeness (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ebin_contamination (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003egtdbtk_species_classification\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ebreadth\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003econANI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003epopANI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eP02AT005BP2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eP02BT0031_bin.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e95.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eEscherichia coli\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eP05AT014BA1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eP05AT0063_bin.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e99.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eEnterococcus faecium\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9997\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9998\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eP14AT043BA1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eP14BT0061_bin.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e97.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9998\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9998\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eP44AT012BA1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eP44AT0283_bin.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e99.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eP55BT045BA1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eP55AT0554_bin.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e98.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.9999\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eHSCT patients in LMICs are highly vulnerable to post-transplant BSIs caused by AMR pathogens. This study builds on a previous investigation by Korula \u003cem\u003eet al.\u003c/em\u003e (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e) in CMC, where bacterial translocation was inferred by matching organisms and AST results between isolates recovered from blood and stool cultures. Here, we used genomics to conduct prospective longitudinal surveillance of stool samples collected from 81 HSCT patients at the same hospital in India, by deploying plate-sweep enrichment to select for targeted priority pathogens, alongside WGS of bloodstream isolates from BSI episodes.\u003c/p\u003e \u003cp\u003eOur findings revealed three key areas for further research and potential clinical impact. First, MacConkey-enriched microbial populations in the HSCT gut were significantly modulated by the pre-transplant conditioning regimen, and their associated resistomes were depleted at the pre-engraftment phase. Wong \u003cem\u003eet al.\u003c/em\u003e reported analogous longitudinal shifts in total faecal microbiome beta-diversities, using 16S data from an Asian cohort undergoing autologous HSCT (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e), which recovered to preconditioning levels within 6-months post-HSCT (beyond our timeframe for follow-up). However, the resistomes in our cohort were promptly restored at early post-engraftment and shifted towards a sustained and significant increase in the abundance of non-efflux AMR genes, including MDR genes, into the late post-engraftment phase. In a longitudinal study with eight paediatric patients in Italy, D\u0026rsquo;Amico \u003cem\u003eet al.\u003c/em\u003e had similar observations with the total gut resistome (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e), reporting an AMR gene profile that had diversified with MDR genes in addition to consolidating the resistome present prior to HSCT. This temporal signature suggests that the early post-engraftment period, when patients are typically recovering and antibiotics administered to counter febrile neutropenia are being de-escalated, paradoxically represents a window of increased AMR risk, which warrants enhanced surveillance to inform antimicrobial stewardship efforts.\u003c/p\u003e \u003cp\u003eHere we generated direct genomic evidence to show that gut colonisation precedes invasive infection in 15.8% of BSI cases in our cohort, establishing proof-of-concept that enrichment-based genomics can identify patients who subsequently developed invasive disease. Blood culture positivity is lower among HSCT patients, since the antimicrobials administered to counter febrile neutropenia during pre-engraftment further reduce the low sensitivity of blood cultures (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). All five patients who had isolates matched between blood and enriched stool samples also showed evidence of enteric mucositis, which has been linked with infections by Gram negative \u003cem\u003eEnterobacteriaceae\u003c/em\u003e, \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e and Gram-positive Enterococci (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Additionally, the enrichment-based strategy offers an alternative mechanism to monitor localised strain transmission of clinically significant pathogens, especially ESBL Enterobacterales (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e), demonstrated by our ability to track isolates with 99.999% ANI longitudinally within and between patients across timepoints. The fact that ~\u0026thinsp;74% of BSI cases could not be linked to the gut highlights the complexity of infection pathogenesis in HSCT recipients and that multiple acquisition routes (nosocomial, respiratory, or gut-associated below detection thresholds) likely contribute.\u003c/p\u003e \u003cp\u003eWe demonstrated that \u0026gt;\u0026thinsp;50% of HSCT recipients carried MDR organisms during the transplant timeline, often harbouring ESBL and carbapenemase genes, which co-occur on diverse plasmids across \u003cem\u003eE. coli\u003c/em\u003e and \u003cem\u003eKlebsiella pneumoniae\u003c/em\u003e species. \u003cem\u003eEnterobacteriaceae\u003c/em\u003e bloom under conditions of intestinal inflammation (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e), which also exacerbates horizontal gene transfer and nosocomial spread in this high-risk population. While WMS hinders the association of AMR genes back to their host species, using plate-sweep enrichment reduces the pool of potential gene hosts compared to WMS, and permits high-resolution screening of targeted bacterial populations compared to isolate-based approaches. Ghosh \u003cem\u003eet al.\u003c/em\u003e reported that the high prevalence of MDR in an Indian HSCT cohort was primarily associated with the \u003cem\u003eblaNDM\u003c/em\u003e and \u003cem\u003eblaOXA\u003c/em\u003e-\u003cem\u003e48-like\u003c/em\u003e carbapenemase genes (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e), but only with pre-transplant surveillance samples. With the longitudinal sampling strategy, we found multiple \u003cem\u003eblaNDM\u003c/em\u003e and \u003cem\u003eblaOXA\u003c/em\u003e genes among the enriched taxa in our cohort, and this abundance significantly increased in the early post-engraftment timepoint when compared to the two preceding timepoints. The microbial clusters we identified suggest possible transmission clusters although environmental surveillance data are required for confirmation.\u003c/p\u003e \u003cp\u003eMethodologically, this study establishes that plate-sweep enrichment offers a practical middle ground between traditional culture, which relies on single colonies and does not provide population or strain-level information, and deep WMS which may be too costly to be conducted at scale to inform clinical decision making. Our data suggest that patients with expanding post-engraftment ESBL/carbapenemase gene abundance could be candidates for modified empirical therapy, pre-emptive decolonisation or enrolment in microbiome restoration trials. However, whether resistome-guided interventions improve outcomes compared to standard care remains an open question requiring randomised controlled trials.\u003c/p\u003e \u003cp\u003eWe did not consider the clinical heterogeneity of HSCT patients. Future investigations could focus on more specific cohorts based on their age, indications for HSCT or a defined outcome, e.g., GI-GvHD. Secondly, selective enrichment can underrepresent slow-growing, fastidious, or anaerobic bacterial populations and by extension, their resistomes. However, our enrichment-based approach was selected to accommodate routine microbiological testing prior to research. Prolonged storage can skew the metagenomic composition of the microbiota in the stool, allowing populations of facultative anaerobes to flourish upon exposure to oxygen if stool collection containers were opened repeatedly, whereas the proportion of obligate anaerobes would stagnate. While our approach limits comparisons with WMS data generated from other HSCT cohorts and detection of rare gut species which could be clinically significant, the MacConkey plate enrichment helped standardise over inconsistencies in sample handling and storage, by selecting key pathogenic members of the microbiota previously identified as important contributors to the burden of AMR infections in HSCT patients.\u003c/p\u003e \u003cp\u003eWhile the enrichment strategy outperforms WMS from a neat sample in capturing intraspecific or population-level heterogeneity for the selected organisms, binning algorithms capture the most dominant STs (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Nonetheless, we were able to account for the population-level heterogeneity across samples because of inStrain\u0026rsquo;s unique \u0026lsquo;microdiversity-aware\u0026rsquo; metrics of conANI and popANI (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Therefore, although we can report bacterial strains co-detected across samples and sample types, we cannot infer transmission source and direction for the remaining\u0026thinsp;~\u0026thinsp;74% of BSI episodes without denser temporal sampling of HSCT patients, more robust spatial tracking, and environmental surveillance.\u003c/p\u003e \u003cp\u003eIn conclusion, we have established the first genomic baseline for AMR dynamics among targeted priority pathogens in HSCT patients in South Asia and provide proof-of-concept that gut reservoirs can be linked to invasive infections. While questions remain about the relative contribution of gut translocation versus other BSI sources, the post-engraftment resistome expansion we observed represents a potentially modifiable risk factor and a clear target for future studies. As HSCT programs expand in LMIC regions with high AMR burden, such surveillance approaches which leverage non-invasive sampling and enrichment-based sequencing will become essential tools for protecting vulnerable transplant populations.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003e \u003cb\u003eFunding and acknowledgements\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThis work was supported by a Wellcome Senior Research Fellowship (215515/Z/19/Z) to Stephen Baker and a grant from the Bill and Melinda Gates Foundation (OPP1159351). The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.\u003c/p\u003e \u003cp\u003eWe would like to thank Agila Kumari Pragasam, Chaitra Shankar, and Ellen Higginson for helpful discussions on the study design. We are also grateful to the media preparation team in the department of Clinical Microbiology at CMC Vellore, and Plamena Naydenova for her assistance with quantifying DNA concentrations. We would also like to thank Jolynne Mokaya for critical feedback on the manuscript. Finally, we would like to acknowledge the clinical teams for their assistance with sample collection and thank the patients and their families for their participation and support.\u003c/p\u003e \u003ch2\u003eConflicts of Interest\u003c/h2\u003e \u003cp\u003eThe authors report there are no competing interests to declare.\u003c/p\u003e \u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConceptualisation and design: AM, SB, BV, SS, BGCollection of samples and clinical data: SS, SD, BGProcessing of samples: AM, SK, YM, DM, PS, KMAnalysis and interpretation of data: AM, JJJ, SBDrafting initial paper: AM, JJJ, SBRevising the final paper: AM, JJJ, SS, SK, SD, YM, DM, PS, KM, BG, BV, SB\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe authors confirm that the data supporting the findings of this study are available in the article and its supplementary materials. The genomic sequencing reads have been deposited in the NCBI Sequence Read Archive under BioProject accession PRJNA1072756.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHeston SM, Young RR, Hong H, Akinboyo IC, Tanaka JS, Martin PL, et al. Microbiology of Bloodstream Infections in Children After Hematopoietic Stem Cell Transplantation: A Single-Center Experience Over Two Decades (1997\u0026ndash;2017). Open Forum Infect Dis. 2020;7(11):ofaa465. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/ofid/ofaa465\u003c/span\u003e\u003cspan address=\"10.1093/ofid/ofaa465\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePouch SM, Satlin MJ. Carbapenem-resistant Enterobacteriaceae in special populations: Solid organ transplant recipients, stem cell transplant recipients, and patients with hematologic malignancies. Virulence. 2017. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/21505594.2016.1213472\u003c/span\u003e\u003cspan address=\"10.1080/21505594.2016.1213472\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e PubMed PMID: 27470662; PubMed Central PMCID: null.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarman P, Choudhary D, Chopra S, Thukral T. Blood stream infections in hematopoietic stem cell transplant patients: A 2-year study from India. Oncol J India. 2020;4(2):43. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4103/oji.oji_7_20\u003c/span\u003e\u003cspan address=\"10.4103/oji.oji_7_20\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStecher B, Denzler R, Maier L, Bernet F, Sanders MJ, Pickard DJ, et al. Gut inflammation can boost horizontal gene transfer between pathogenic and commensal Enterobacteriaceae. Proc Natl Acad Sci. 2012;109(4):1269\u0026ndash;74. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.1113246109\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1113246109\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnthony WE, Burnham CAD, Dantas G, Kwon JH. The Gut Microbiome as a Reservoir for Antimicrobial Resistance. J Infect Dis. 2021;223(Supplement_3):S209\u0026ndash;13. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/infdis/jiaa497\u003c/span\u003e\u003cspan address=\"10.1093/infdis/jiaa497\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSardzikova S, Andrijkova K, Svec P, Beke G, Klucar L, Minarik G, et al. Gut diversity and the resistome as biomarkers of febrile neutropenia outcome in paediatric oncology patients undergoing hematopoietic stem cell transplantation. Sci Rep. 2024;14(1):5504. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-024-56242-8\u003c/span\u003e\u003cspan address=\"10.1038/s41598-024-56242-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIskandar K, Molinier L, Hallit S, Sartelli M, Hardcastle TC, Haque M, et al. Surveillance of antimicrobial resistance in low- and middle-income countries: a scattered picture. Antimicrob Resist Infect Control. 2021;10(1):63. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13756-021-00931-w\u003c/span\u003e\u003cspan address=\"10.1186/s13756-021-00931-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhazal SS, Stevens MP, Bearman GM, Edmond MB. Utility of surveillance blood cultures in patients undergoing hematopoietic stem cell transplantation. Antimicrob Resist Infect Control. 2014;3:20. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/2047-2994-3-20\u003c/span\u003e\u003cspan address=\"10.1186/2047-2994-3-20\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e PubMed PMID: 24999384; PubMed Central PMCID: PMC4082284.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurray CJL, Ikuta KS, Sharara F, Swetschinski L, Robles Aguilar G, Gray A, et al. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. The Lancet. 2022;399(10325):629\u0026ndash;55. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(21)02724-0\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(21)02724-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKrakau S, Straub D, Gourl\u0026eacute; H, Gabernet G, Nahnsen S. nf-core/mag: a best-practice pipeline for metagenome hybrid assembly and binning. NAR Genomics Bioinforma. 2022;4(1):lqac007. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nargab/lqac007\u003c/span\u003e\u003cspan address=\"10.1093/nargab/lqac007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGurevich A, Saveliev V, Vyahhi N, Tesler G. QUAST: quality assessment tool for genome assemblies. Bioinformatics. 2013;29(8):1072\u0026ndash;5. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btt086\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btt086\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUritskiy GV, DiRuggiero J, Taylor J. MetaWRAP\u0026mdash;a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome. 2018;6(1):158. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s40168-018-0541-1\u003c/span\u003e\u003cspan address=\"10.1186/s40168-018-0541-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang DD, Li F, Kirton E, Thomas A, Egan R, An H, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ. 2019;7:e7359. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7717/peerj.7359\u003c/span\u003e\u003cspan address=\"10.7717/peerj.7359\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu YW, Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics. 2016;32(4):605\u0026ndash;7. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btv638\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btv638\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlneberg J, Bjarnason BS, De Bruijn I, Schirmer M, Quick J, Ijaz UZ, et al. Binning metagenomic contigs by coverage and composition. Nat Methods. 2014;11(11):1144\u0026ndash;6. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nmeth.3103\u003c/span\u003e\u003cspan address=\"10.1038/nmeth.3103\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChaumeil PA, Mussig AJ, Hugenholtz P, Parks DH. GTDB-Tk: a toolkit to classify genomes with the Genome Taxonomy Database. Hancock J, editor. Bioinformatics. 2020;36(6):1925\u0026ndash;7. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btz848\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btz848\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome Biol. 2019;20(1):257. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13059-019-1891-0\u003c/span\u003e\u003cspan address=\"10.1186/s13059-019-1891-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu J, Breitwieser FP, Thielen P, Salzberg SL. Bracken: estimating species abundance in metagenomics data. PeerJ Comput Sci. 2017;3:e104. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7717/peerj-cs.104\u003c/span\u003e\u003cspan address=\"10.7717/peerj-cs.104\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu J, Rincon N, Wood DE, Breitwieser FP, Pockrandt C, Langmead B, et al. Metagenome analysis using the Kraken software suite. Nat Protoc. 2022;17(12):2815\u0026ndash;39. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41596-022-00738-y\u003c/span\u003e\u003cspan address=\"10.1038/s41596-022-00738-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFoster ZSL, Sharpton TJ, Gr\u0026uuml;nwald NJ. Metacoder: An R package for visualization and manipulation of community taxonomic diversity data. Poisot T, editor. PLOS Comput Biol. 2017;13(2):e1005404. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pcbi.1005404\u003c/span\u003e\u003cspan address=\"10.1371/journal.pcbi.1005404\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOksanen J, Simpson GL, Blanchet FG, Kindt R, Legendre P, Minchin PR, et al. vegan: Community Ecology Package [Internet]. 2001 [cited 2025 Nov 21]. p. 2.7-2. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://CRAN.R-project.org/package=vegan\u003c/span\u003e\u003cspan address=\"https://CRAN.R-project.org/package=vegan\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.32614/CRAN.package.vegan\u003c/span\u003e\u003cspan address=\"10.32614/CRAN.package.vegan\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArbizu PM. pairwiseAdonis: pairwise multilevel comparison using adonis. 2017. 2019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOlm MR, Crits-Christoph A, Bouma-Gregson K, Firek BA, Morowitz MJ, Banfield JF. inStrain profiles population microdiversity from metagenomic data and sensitively detects shared microbial strains. Nat Biotechnol. 2021;39(6):727\u0026ndash;36. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41587-020-00797-0\u003c/span\u003e\u003cspan address=\"10.1038/s41587-020-00797-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCsardi MG. Package \u0026lsquo;igraph\u0026rsquo;. Last Accessed. 2013;3(09):2013.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePedersen TL. tidygraph: A Tidy API for Graph Manipulation [Internet]. 2023. Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/thomasp85/tidygraph\u003c/span\u003e\u003cspan address=\"https://github.com/thomasp85/tidygraph\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSi B, Liang Y, Zhao J, Zhang Y, Liao X, Jin H, et al. GGraph: An Efficient Structure-Aware Approach for Iterative Graph Processing. IEEE Trans Big Data. 2022;8(5):1182\u0026ndash;94. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TBDATA.2020.3019641\u003c/span\u003e\u003cspan address=\"10.1109/TBDATA.2020.3019641\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorula A, Perumalla S, Devasia AJ, Abubacker FN, Lakshmi KM, Abraham A, et al. Drug-resistant organisms are common in fecal surveillance cultures, predict bacteremia and correlate with poorer outcomes in patients undergoing allogeneic stem cell transplants. Transpl Infect Dis. 2020;22(3). doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/tid.13273\u003c/span\u003e\u003cspan address=\"10.1111/tid.13273\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWong SP, Er YX, Tan SM, Lee SC, Rajasuriar R, Lim YAL. Oral and Gut Microbiota Dysbiosis is Associated with Mucositis Severity in Autologous Hematopoietic Stem Cell Transplantation: Evidence from an Asian Population. Transplant Cell Ther. 2023;29(10):633.e1-633.e13. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jtct.2023.06.016\u003c/span\u003e\u003cspan address=\"10.1016/j.jtct.2023.06.016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eD\u0026rsquo;Amico F, Soverini M, Zama D, Consolandi C, Severgnini M, Prete A, et al. Gut resistome plasticity in pediatric patients undergoing hematopoietic stem cell transplantation. Sci Rep. 2019;9(1):5649. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-019-42222-w\u003c/span\u003e\u003cspan address=\"10.1038/s41598-019-42222-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNieman AE, Savelkoul PHM, Beishuizen A, Henrich B, Lamik B, MacKenzie CR, et al. A prospective multicenter evaluation of direct molecular detection of blood stream infection from a clinical perspective. BMC Infect Dis. 2016;16(1):314. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12879-016-1646-4\u003c/span\u003e\u003cspan address=\"10.1186/s12879-016-1646-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBalletto E, Mikulska M. Bacterial infections in haematopoietic stem cell transplant patients. Mediterr J Hematol Infect Dis. 2015;7:e2015045. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4084/mjhid.2015.045\u003c/span\u003e\u003cspan address=\"10.4084/mjhid.2015.045\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJazmati T, Hamprecht A, Jazmati N. Comparison of stool samples and rectal swabs with and without pre-enrichment for the detection of third-generation cephalosporin-resistant Enterobacterales (3GCREB). Eur J Clin Microbiol Infect Dis. 2021;40(11):2431\u0026ndash;6. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10096-021-04250-1\u003c/span\u003e\u003cspan address=\"10.1007/s10096-021-04250-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhosh S, Bhattacharya S, Goel G, Deshmukh RA, Javed R, Roychowdhury M, et al. Hematopoietic stem-cell transplantation in a zoo of multidrug‐resistant organisms: Data from a cancer center in eastern India. Transpl Infect Dis. 2023;e14072. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/tid.14072\u003c/span\u003e\u003cspan address=\"10.1111/tid.14072\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9289820/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9289820/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003ePatients undergoing haematopoietic stem cell transplantation (HSCT) are highly susceptible to bloodstream infections (BSIs) caused by antimicrobial resistant (AMR) pathogens, yet the temporal dynamics and clinical relevance of intestinal AMR reservoirs remain poorly defined, particularly in high-burden settings.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe conducted prospective longitudinal surveillance of 81 HSCT recipients in India using targeted enrichment-based stool metagenomics, coupled with whole genome sequencing of bloodstream isolates and strain-level tracking.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eAcross 252 samples, we identified a restructuring of the gut-associated resistome, characterised by depletion during conditioning and a marked post-engraftment expansion of non-efflux AMR determinants, including plasmid-borne carbapenemase and ESBL genes (\u003cem\u003ebla\u003c/em\u003eNDM, \u003cem\u003ebla\u003c/em\u003eOXA). This expansion defined a previously underappreciated window of vulnerability during early immune recovery. Over 50% of patients harboured multidrug-resistant organisms, and carriage of \u003cem\u003ebla\u003c/em\u003eNDM was associated with increased odds of subsequent infection. Strain-resolved analyses provided genomic evidence linking gut colonisation to bloodstream infection in a subset of cases, demonstrating the feasibility of non-invasive prediction of invasive disease. Patients with AMR-BSIs experienced high mortality (\u0026gt;\u0026thinsp;40%).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThese findings establish that dynamic changes in the gut resistome following HSCT can identify periods of heightened infection risk and provide a foundation for predictive, genomics-guided surveillance and antimicrobial stewardship strategies in high-burden settings.\u003c/p\u003e","manuscriptTitle":"Stool-based genomic surveillance identifies post-engraftment expansion of multidrug-resistant pathogens in haematopoietic stem cell transplant patients in India","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-26 17:10:33","doi":"10.21203/rs.3.rs-9289820/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d58d216b-4d02-4d3d-8973-5ea757ebd219","owner":[],"postedDate":"April 26th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Rejected","date":"2026-05-11T18:17:31+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-07T06:02:21+00:00","index":28,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-05T11:40:25+00:00","index":26,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":66859936,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":66859937,"name":"Health sciences/Diseases"},{"id":66859938,"name":"Biological sciences/Genetics"},{"id":66859939,"name":"Biological sciences/Microbiology"}],"tags":[],"updatedAt":"2026-05-11T18:25:42+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-26 17:10:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9289820","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9289820","identity":"rs-9289820","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}