Microbiology, Antibiotic Resistance, and Antimicrobial Burden in Infected Necrotizing Pancreatitis: A Single-Center Cohort

Authors:
Ioannis Karikis, Jack H. A. Miller, Arielle M. Moore, Peter J. Fagenholz, Yasmin G. Hernandez-Barco, Miriam B. Barshak, Casey M. Luckhurst

Body of Abstract:
Background:
Infected necrotizing pancreatitis (INP) can be associated with prolonged critical illness and extensive antimicrobial exposure. Contemporary microbiology and resistance patterns are poorly characterized. We aimed to describe the microbiology, organism-specific resistance, and antimicrobial burden of INP.

Methods:
We conducted a retrospective cohort study of adult patients with necrotizing pancreatitis admitted to a tertiary academic hospital between 2019 and 2025 who underwent invasive intervention on pancreatic and/or peripancreatic necrosis. Infection prompting intervention was defined as the presence of gas on computed tomography and/or clinical deterioration; only cultures obtained at the time of the index invasive procedure were analyzed. We described baseline characteristics, microbiology, patient-level antibiotic resistance, and antimicrobial use patterns, focusing on empiric regimens at first antibiotic exposure and antifungal therapy.

Results:
Among 48 patients with INP, mean age was 55.5 ± 16.5 years and median hospital length of stay was 27.5 days (IQR 17.5–67.5); 54.2% required ICU admission and median ICU length of stay was 16.0 days (IQR 8.0–30.0). Evidence of infection prior to intervention included gas on CT in 22/48 (45.8%) and clinical deterioration in 26/48 (54.2%). Bacteria were isolated in 44/48 patients (91.7%) and fungal species in 11/48 (22.9%). The most common organisms were Escherichia coli (29.2%), Enterococcus faecium (22.9%), Candida albicans (14.6%), Enterococcus faecalis (14.6%), and Staphylococcus epidermidis (14.6%); 70.5% of infections were polymicrobial. 34/44 (77.3%) had at least one antibiotic-resistant isolate. Empiric regimens at first antibiotic exposure most commonly included piperacillin–tazobactam (24/48, 50.0%), meropenem (11/48, 22.9%), vancomycin (9/48, 18.8%), metronidazole (8/48, 16.7%), and cefepime (7/48, 14.6%). Among patients who received these agents empirically and had susceptibility data, resistant isolates were identified in 5/22 (22.7%) for piperacillin–tazobactam, 1/10 (10.0%) for meropenem, 0/6 (0%) for cefepime, 2/8 (25.0%) for vancomycin. Despite fungal isolation in only 11/48 (22.9%), 31/48 (64.6%) received at least one antifungal agent, most commonly micafungin (41.7%) and fluconazole (25.0%).

Conclusions:
Infected necrotizing pancreatitis is characterized by polymicrobial collections with both gram-positive and gram-negative organisms and frequent antibiotic resistance. These findings highlight INP as a major target for antimicrobial stewardship and suggest that organism-specific resistance data should inform future empiric antimicrobial and antifungal strategies.