2409. Drug-Induced Liver Injury (DILI) in a National Cohort of Hospitalized Patients Treated With Aztreonam and Ceftazidime

Abstract

BackgroundDILI, although uncommon, can be a severe and even fatal complication of antibiotic use. The safety of novel regimens targeting MDR Gram-negative bacteria (GNB) is an important concern. Cephalosporins such as ceftazidime (CAZ) are rare causes of clinically apparent DILI, while data regarding DILI with other antibiotics such as the monobactam aztreonam (ATM) are sparse. ATM and CAZ are partnered with many novel β-lactamase inhibitors (i.e., avibactam, AVI) as therapy for MDR infections (CAZ-AVI and ATM-AVI) We aimed to compare the incidence and type of DILI associated with ATM and CAZ. MethodsUsing a cohort of patients hospitalized within Veterans Health Administration (VHA), we identified patients treated with ATM or CAZ for 3 or more consecutive days who also had LFTs measured during (day 3 or later) or within 7 days of stopping treatment. We excluded patients with abnormal LFTs in the year prior to ATM or CAZ treatment. Using alanine aminotransferase, alkaline phosphatase, and bilirubin measures, we applied clinical chemistry criteria to identify cases of DILI. We applied further criteria to classify DILI according to clinical pattern and severity (mild vs. moderate/severe), comparing the relative frequencies between ATM and CAZ.ResultsAmong 18,813 courses of CAZ or ATM, 3,432 ATM and 2,662 CAZ courses met our criteria (Figure 1). While the overall rate of any DILI was higher in ATM than CAZ (5.8% vs. 3.2%, P < 0.01), the rate of moderate/severe DILI was similarly low for both agents (1.6% in ATM vs. 1.3% in CAZ, P = 0.3). The clinical pattern of DILI cases differed by drug, with the hepatocellular pattern comprising a larger proportion of the ATM DILI cases (37%) than the CAZ DILI cases (25%) and the cholestatic pattern comprising a smaller proportion (48% vs. 61%) (Figure 2).ConclusionIn this national cohort of hospitalized patients treated with ATM or CAZ, the overall rate of DILI was significantly higher in patients treated with ATM than in those treated with CAZ. However, there is a similarly low rate of moderate/severe DILI. Although further analyses are required to better understand causal mechanisms and clinical risks of DILI in patients receiving ATM or CAZ, these data from a large national cohort provide a useful benchmark of drug safety. Disclosures T. Lodise, paratek: Consultant and Scientific Advisor, Consulting fee.