Outcomes of treating AmpC-producing Enterobacterales bacteraemia with carbapenems vs. non-carbapenems.

INTRODUCTION: AmpC ß-lactamases are found in Enterobacter species, Serratia species, Citrobacter freundii, Providencia species and Morganella morganii ('ESCPM'). Carbapenems are commonly used to treat severe 'ESCPM' infections. Carbapenem-sparing agents are needed because of increasing carbapenem resistance worldwide. Use of cefepime and piperacillin-tazobactam has limited supportive clinical data. We evaluated the efficacy of non-carbapenems vs. carbapenems in 'ESCPM' bacteraemia. METHODS: A retrospective cohort study was conducted on patients with 'ESCPM' bacteraemia. Primary outcome was 30-day mortality. Analyses were performed on patients who received carbapenems vs. piperacillin-tazobactam or cefepime monotherapy as empirical and definitive therapy. Propensity score for carbapenem therapy was adjusted for in multivariate analyses for 30-day mortality. RESULTS: A total of 241 patients were included. The most common bacterium isolated was Enterobacter species (58.1%). Common sources were urinary (22.8%) and vascular lines (22.0%). Carbapenems (28.6%) and piperacillin-tazobactam (28.6%) were the commonest empirical antibiotics. Carbapenems (54.8%) and cefepime (23.7%) were the most common definitive antibiotics. Median Pitt bacteraemia score was 1 (interquartile range [IQR], 0-2). Overall, 30-day mortality was 12.9%. Adjusted multivariate analyses for empirical and definitive antibiotic treatment models yielded risk factors for 30-day mortality, including higher Pitt bacteraemia score (empirical: adjusted OR [aOR] 1.21 for each point increase, 95% confidence interval [CI]:1.01-1.45; definitive: aOR 1.33 for each point increase, 95% CI:1.06-1.69) and age (empirical: aOR 1.04 for each year increase, 95% CI:1.01-1.08). Empirical piperacillin-tazobactam (aOR 0.29, 95% CI:0.07-1.27) and definitive cefepime (aOR 0.65, 95% CI:0.12-3.55) were not associated with 30-day mortality. CONCLUSIONS: Compared with carbapenem therapy, empirical piperacillin-tazobactam and definitive cefepime were not associated with 30-day mortality in 'ESCPM' bacteraemia.

Nuclear import of the varicella-zoster virus latency-associated protein ORF63 in primary neurons requires expression of the lytic protein ORF61 and occurs in a proteasome-dependent manner.

Varicella-zoster virus (VZV) open reading frame (ORF) 63 protein (ORF63p) is one of six VZV ORFs shown to be transcribed and translated in latently infected human dorsal root ganglia. ORF63p accumulates exclusively in the cytoplasm of latently infected sensory neurons, whereas it is both nuclear and cytoplasmic during lytic infection and following reactivation from latency. Here, we demonstrate that infection of primary guinea pig enteric neurons (EN) with an adenovirus expressing ORF63p results in the exclusive cytoplasmic localization of the protein reminiscent of its distribution during latent VZV infection in humans. We show that the addition of the simian virus 40 large-T-antigen nuclear localization signal (NLS) results in the nuclear import of ORF63p in EN and that the ORF63p endogenous NLSs are functional in EN when fused to a heterologous protein. These data suggest that the cytoplasmic localization of ORF63p in EN results from the masking of the NLSs, thus blocking nuclear import. However, the coexpression of ORF61p, a strictly lytic VZV protein, and ORF63p in EN results in the nuclear import of ORF63p in a proteasome-dependent manner, and both ORF63p NLSs are required for this event. We propose that the cytoplasmic localization of ORF63p in neurons results from NLS masking and that the expression of ORF61p removes this block, allowing nuclear import to proceed.