Microbial Signatures of Thrombosis and Organ Failure in Surgical ICU Patients

Authors:
Ioannis Karikis, Yasmin Arda, Galit H. Frydman

Body of Abstract:
Background:
Surgical ICU patients, exposed to severe physiological stress and broad-spectrum antimicrobials, may be vulnerable to pathogen-driven thrombosis and organ injury, but species-level data are limited. We aimed to determine whether specific species-level microbial patterns are enriched in SICU patients who develop major thrombosis and/or organ failure.

Methods:
We conducted a single-center retrospective cohort study of 137 surgical ICU (SICU) patients with suspected infection and available culture and sensitivity data who underwent longitudinal microbiologic sampling during their SICU admission between 2020 and 2024. For each patient, we identified growth of bacterial and fungal species, including speciation and antibiotic sensitivity data, as well as concurrent viral infection, as clinically indicated. The primary outcome was a major thrombotic event (MTE) (deep venous thrombosis, pulmonary embolism, line thrombosis, myocardial infarction, or stroke). Secondary outcomes were acute kidney injury (AKI) and acute liver failure (ALF) during the index hospitalization. Associations between microbial patterns and outcomes were evaluated descriptively using χ² or Fisher’s exact tests, as appropriate.

Results:
Of 137 patients, 22 (16%) developed an MTE. Among 132 patients with available renal and liver data, 41 (31%) developed AKI and 10 (8%) developed ALF. Thrombosis was more common in patients with Candida and several bacterial pathogens. Any Candida species were present in 36% of patients with thrombosis versus 4% without (p<0.001), driven predominantly by non-albicans Candida. Thrombotic patients were also more likely to grow Enterococcus species (23% vs 6%, p=0.038), Escherichia coli (E. coli) (27% vs 5%, p=0.003), Staphylococcus aureus (S. aureus) (14% vs 2%, p=0.023), and showed a trend towards Pseudomonas species (18% vs 4%, p=0.052). AKI exhibited a similar microbial signature. Candida species were again enriched in patients with AKI (27% vs 2%, p<0.001), as were Klebsiella pneumoniae (K. pneumoniae) (17% vs 6%, p=0.032), Enterococcus species (20% vs 4%, p=0.005), and E. coli (17% vs 6%, p=0.032). AKI also clustered with COVID-19 infection (78% vs 48%, p=0.001). ALF was less frequent but showed signals for Enterococcus faecium (E. faecium) (30% vs 7%, p=0.017) and K. pneumoniae (20% vs 2%, p=0.001) in a small subgroup. Conclusions: In this SICU cohort, MTE, AKI, and ALF each appeared to have distinct infectious-agent signatures. These findings support the hypothesis that specific pathogens may play a significant role in the development of thrombosis and multi-organ failure in critically ill patients. These exploratory data are hypothesis-generating and justify larger studies that integrate microbiology, coagulation, and organ-dysfunction endpoints in the SICU population.