Utility of paired plasma and drainage fluid mNGS in diagnosing acute intra-abdominal infections with sepsis

This single-center prospective observational study was conducted in the ICU (Intensive Care Unit) department of the Peking Union Medical College Hospital (PUMCH), which is a tertiary academic hospital. Septic patients with severe acute IAI requiring ICU care from the emergency department were admitted from October 2021 to December 2022. Acute IAI were diagnosed according to previous guidelines [12‐14], patients with source control procedures like surgery or percutaneous puncture drainage undergone on the first day of admission were enrolled in this study. The exclusion criteria were as following: (1)<18-year-old; (2) ICU stay ≤ 24 h; (3) incomplete clinical data, like missing data such as admission diagnosis, antimicrobial therapy records, without access to drainage fluid, etc.; (4) refusal to undergo mNGS detection; and (5) failure to obtain written consent. The ethics committee of PUMCH approved this study (approval number: JS-1170). Informed consent was obtained from the next kin of all patients, and the study was registered with the Chinese Clinical Trial Registry (identifier ChiCTR-ROC-17,010,750).

In addition, to investigate the influence of mNGS on clinical treatment, a retrospective cohort (non-mNGS group) was included for comparison with the prospective mNGS group. The non-mNGS group comprised patients with acute IAIs admitted to our department from August 2020 to October 2021, before the standardized implementation of mNGS testing for peritoneal drainage fluid samples at our center. All patients in both groups met the same inclusion and exclusion criteria and received empiric antibiotic therapy according to relevant guidelines. The comparison between the two groups aimed to assess the impact of mNGS results on antibiotic utilization patterns.

Initial empirical medications in both these two groups were administered according to relevant IAI guidelines [15, 16]. Empiric antibiotic therapy should include agents against aerobic Gram-negative bacteria (e.g., Enterobacteriaceae), aerobic streptococci, and obligate enteric anaerobic organisms. Resistant or opportunistic pathogens such as Candida spp. should also be covered in selected conditions like a healthcare setting, corticosteroid use, organ transplantation or previous antimicrobial therapy. Antibiotic therapy was also selected based on based on infection`s severity, community or hospital acquisition, presence of septic shock, and the degree of peritoneal observed during surgery [12, 17, 18]. Empiric antimicrobial therapy should be narrowed once culture and susceptibility results are available and adequate clinical improvement is noted. In mNGS group, the therapeutic regimens were adjusted based on mNGS detection, such as lack of evidence of certain bacteria may indicate early discontinuation of corresponding treatment. The retrospective non-mNGS group collected septic patients with acute IAI admitted to our department from August 2020 to October 2021. The inclusion and exclusion criteria were consistent with the above. Only CMTs of plasma and PD was sent for pathogen detection. All enrolled patients underwent the same diagnosis and treatment procedure in our study.

Clinical history evaluation and laboratory tests were carried out, including age, sex, routine blood examination, procalcitonin (PCT), hemodynamic parameters, respiratory parameters, Sequential Organ Failure Assessment (SOFA) score, Acute Physiology and Chronic Health Evaluation (APACHE) II score and experiential antibiotic use upon ICU admission. Duration of ICU stay time and 28-day mortality rate were recorded as follow-up data. Peripheral blood samples and intraperitoneal fluid samples were obtained for mNGS analysis and culture-based microbiological tests immediately after source control procedures. At least 5mL of blood sample and PD fluid was collected for adequate microbiological testing, and samples were sent to laboratory immediately avoiding influence on results.

According to procedures of the central laboratory of the Clinical Laboratory Department PUMCH, blood and PD fluid culture and confirmation of species identification were performed. For blood culture, standard aerobic and anaerobic bottles were processed by the microbiology laboratory according to standard protocol using Bac T Alert,5 a continuously monitored, carbon dioxide detection system. Specimens were incubated for 7 days and all positive vials were inoculated onto appropriate media and further processed by culture-based techniques [19]. For PD fluid culture, the sample was performed using routine isolation media, including blood agar, eosin methylene-blue (EMB) agar, Mueller-Hinton agar and cooked meat medium (Oxoid, UK). Plates were incubated at 37 °C in 5–10% CO2 for 24–48 h using BC60 automated culture system (Autobio Diagnostics Co., Ltd., China). The strains were then isolated and identified the species using the VITEK-2 Compact Instrument (bioMérieux, France).

Plasma and PD fluid of each patient were obtained to perform metagenomic next-generation sequencing and data analysis. The mNGS detection process includes host cell removal, nucleic acid extraction, library preparation, sequencing, and bioinformatics analysis. Briefly, DNA was extracted using a DNA extraction kit (Tiangen Biotech (Beijing) Co., Ltd., China) in a 1 mL sample. cDNA was generated using reverse transcriptase and dNTPs (Thermo Fisher). Libraries were constructed for the DNA and cDNA samples using a Nextera XT DNA Library Prep Kit (Illumina, San Diego, CA). Library pools were loaded onto the Illumina Nextseq CN500 sequencer for 75 cycles of single-end sequencing, generating approximately 20 to 40 million reads for each library. Quality control and evaluation of FASTQ format data obtained by sequencing were carried out, and low-quality or undetected sequences, sequences contaminated by splices, high-coverage repeats, and short read-length sequences were filtered to retain high-quality sequencing data. Microorganism identification was obtained through mapping to the commercial pathogen database as previously described [20]. In the mNGS analysis, the relative abundances of microbial taxa were determined and compared to predefined abundance cut-offs based on previous studies to differentiate likely pathogens from commensals/contaminants. Low-abundance organisms were excluded from analysis. The sequencing data are available at the NCBI SRA database via BioProject accession PRJNA987137.

The primary outcome was the difference in microbiologic positivity rates between mNGS and conventional cultures for PD fluid and plasma. Secondary outcomes included the concordance between plasma and PD mNGS results and their correlation with clinical parameters, the microbiota profiles in intra-abdominal infections, and the impact of mNGS on empirical antimicrobial therapy in terms of agent selection and duration.