Authors: Shannon R. Mayberry, Cristina Martinez, Benjamin Albrecht, Sujit Suchindran, Sarah B. Green
Resident e-mail for contact:
[email protected] Standard of Care compared to alternative beta-lactam agents in high-inoculum AmpC infections
Purpose/Background:
Treatment of low-risk AmpC-inducible organisms should rely on antimicrobial susceptibility testing (AST) to guide treatment of infection. Current Infectious Diseases Society of America (IDSA) guidelines identify Serratia marcesens, Morganella morganii, and Providencia spp. as low-risk species. Per this guidance, when susceptibility to agents such as ceftriaxone and piperacillin-tazobactam is demonstrated, treatment with that antibiotic is considered appropriate. However, infections with high bacterial burden, including endocarditis and central nervous system (CNS) infections, pose an additional challenge for optimal treatment selection. The IDSA guidelines state cefepime may be reasonable to utilize for high-inoculum infections despite susceptibility to ceftriaxone, due to less risk of AmpC hydrolysis. However, efficacy to cefepime in this setting is uncertain given potential for an inoculum effect, while carbapenems appear less affected. Clinical evidence remains limited, and treatment guidance for high-inoculum AmpC infections is unclear. In this retrospective case series, we describe the clinical outcomes of patients treated with a carbapenem, cefepime, or an alternative beta-lactam agent for deep-seated endovascular and CNS infections due to low-risk AmpC-inducible organisms.
Methods:
This IRB approved, retrospective case series included adults (≥18 years) admitted to Emory Healthcare acute care hospitals between January 1, 2023, and January 31, 2025, with endocarditis, endovascular, or CNS infection due to a low-risk AmpC-inducible organism (Serratia marcescens, Morganella morganii, or Providencia spp.). Patients were identified through diagnosis codes (e.g. endocarditis, endovascular infection, CNS infection) and a positive blood or CNS culture for a low-risk AmpC organism. A total of 28 patients were identified for review. Further exclusion criteria were polymicrobial cultures, switch of therapy after 72 hours, or definitive non–β-lactam or combination therapy. Of this, 10 patients were included for final evaluation. Case data collected included patient demographics, comorbidities, infection characteristics, microbiology, antimicrobial therapy, 30-day all-cause mortality, 90-day infection recurrence, 90-day readmission, and adverse events.
Results:
Ten patients with high-inoculum infections due to low-risk, high inoculum AmpC-inducible organisms were included. Nine infections were caused by Serratia marcescens and one by Morganella morganii. Infection types included endovascular infections (n=6), central nervous system infections (n=3), and infective endocarditis (n=1). Initial antimicrobial regimens varied and included ceftriaxone, cefepime, ceftazidime, and piperacillin/tazobactam, with several patients undergoing antibiotic escalation after organism identification. All but one patient had infectious diseases consultation. Antibiotic modifications were common, ranging from zero to eight changes during admission. One patient experienced inpatient mortality. A single patient had a 90-day infection-related readmission. No consistent signal of clinical failure was observed among patients treated with ceftriaxone compared with cefepime or carbapenems. Despite frequent antimicrobial adjustments, most patients achieved clinical stability without recurrent infection or excess mortality within 90 days.
Conclusions:
In this small retrospective case series, clinical outcomes were similar among patients receiving ceftriaxone, cefepime, or carbapenems for high-inoculum infections caused by low-risk AmpC-inducible organisms. Rates of inpatient mortality and 90-day readmission were low, and no clear outcome advantage was observed with broader-spectrum therapy. These findings suggest that carefully selected non-carbapenem β-lactams may be a reasonable option in certain deep-seated infections when supported by susceptibility data and close clinical monitoring. However, frequent antibiotic escalation highlights ongoing clinician concern regarding inducible resistance and the inoculum effect. Given the limited sample size and descriptive design, definitive comparative conclusions cannot be drawn. Larger, multicenter studies are needed to better define optimal therapy in this high-risk population. Until additional data are available, antimicrobial stewardship decisions should balance theoretical resistance risks with the goal of preserving carbapenem activity.
