Comparison of the effect of carbapenem and piperacillin-tazobactam on the development of fecal colonisation with carbapenem resistant enterobacterales

Carbapenems are one of the most important treatment options for infections caused by multidrug-resistant bacteria. The incidence of carbapenem-resistant Enterobacterales (CRE) is increasing, which makes managing patients infected with these bacteria challenging. Morbidity and mortality rates are significantly on the rise with limited treatment options for CRE infections. Prolonged hospitalization, immunosuppression, invasive procedures, treatment in the intensive care unit, and antibiotic use are well-known risk factors for colonization and infection with CRE. Due to their broad spectrum of action, piperacillin-tazobactam (TZP) and carbapenems are frequently preferred antibiotics for empirical treatment in critically ill patients. However, both carbapenem and TZP use have been reported as risk factors for the development of CRE infections. Determining which antibiotic is associated with a higher risk of developing CRE infection is difficult because both antibiotics can be used consecutively during hospitalization and patients have other risk factors for CRE infection. This study aims to investigate the risk of developing CRE colonization in an experimental setting with no risk factors other than antibiotic use. This study was conducted at the Department of Infectious Diseases and Clinical Microbiology Laboratory and the Department of Pharmacology Laboratory at Hacettepe University Faculty of Medicine and Faculty of Pharmacy, respectively. In the first stage of the two-step study, groups which included six mice each were divided into treatment groups receiving TZP, meropenem, cyclophosphamide, TZP and cyclophosphamide, or meropenem and cyclophosphamide, as well as a control group. A total of 36 mice were used in six different groups. Antibiotics were administered for 14 days, and fresh stool samples were collected from the mice on days zero, three, five, seven, ten, and fourteen. The samples were kept in liquid media supplemented with meropenem and vancomycin discs for 72 hours, and turbidity was checked daily. In the case of turbidity, a double-sided solid medium was inoculated. The bacteria were examined microscopically by gram staining, which revealed growth. Bacteria with gram-negative bacilli or coccobacilli morphology were identified by MALDI-TOF. Antibiotic susceptibility was investigated using the disc diffusion method. On the fourteenth day of the experiments, real-time polymerase chain reaction (RT-PCR) was performed to determine the genes causing carbapenem resistance. In the second stage of the study, three different strains ERT-1: E. coli (susceptible to all antibiotics); ERT-3: K. pneumoniae (susceptible to all antibiotics tested except ampicillin); ERT-4: Extended-spectrum beta-lactamase (ESBL) producing K. pneumoniae were used. ERT-1 and ERT-3 isolates were mixed with blaOXA-48 gene-positive isolates of the same species isolated from patient clinical specimens to create a heterogeneous bacterial population. New E. coli and K. pneumoniae strains (ERT-2 and ERT-5, respectively) were added to the study, and in vitro experiment was initiated. Minimum inhibitory concentration (MIC) values of bacteria were determined by microdilution method using ertapenem and TZP antibiotics. For each bacterium in the microplate, a double-sided solid medium was inoculated from the last well in which growth was observed and incubated for 24 hours and microdilution was repeated on the next day. Two TZP susceptible isolates (TZP-1: E. coli susceptible to all antibiotics; TZP-3: K. pneumoniae susceptible to all tested antibiotics except ampicillin) were exposed to TZP by microdilution method in a different microplate. After two consecutive TZP exposures, ertapenem MIC values were monitored. The experiment was completed when the bacteria developed resistance after ertapenem exposure. For strains that remained susceptible to ertapenem despite antibiotic exposure, the experiments were terminated after ten consecutive antibiotic exposures. It was planned to conduct a one-way analysis of variance followed by Tukey's test for group comparisons, to present the results as mean±standard error, and to accept P< 0.05 as statistically significant.. The study was started with 36 mice. Enterobacterales were isolated in 17 (8.6%) of 196 stool cultures obtained from mice during the experiments. When the Enterobacterales grown in fecal samples from each experimental animal were identified to species level, ten bacteria were identified as K. pneumoniae, two as E. coli, two as M. morganii, two as E. cloacae complex and one as K. oxytoca. All of the Enterobacterales isolated from stool cultures were susceptible to carbapenem. In the meropenem-treated group (n=6) and the control group (n=6), no Enterobacterales were grown in any of the stool cultures. In stool cultures obtained from TZP-treated mice (n=6), ESBL-producing Enterobacterales were isolated in three of them on days 7, 10, and 14 of antibiotic administration. In the group treated with TZP and cyclophosphamide (n=6), ESBL-producing Enterobacterales were observed in a total of five samples; one on day 7, one on day 10, and three on day 14 of TZP administration. On day 14 of the experiments, RT-PCR assays were performed on stool samples collected from each mice. One sample from the cyclophosphamide-treated group was positive for the blaNDM gene, but no Enterobacterales were grown in the stool culture. As no CRE growth was detected in any subject and ESBL-producing Enterobacteriales were found only in the TZP-treated group, statistical analyses were not performed. In in vitro experiments, strain ERT-1 was still susceptible to ertapenem after ten consecutive exposures to ertapenem. However, the new E. coli strain (ERT-2), obtained by mixing this strain with another E. coli strain producing OXA-48 carbapenemases, became resistant to ertapenem after the fourth exposure to ertapenem. The K. pneumoniae strain encoded by ERT-3 the ertapenem MIC value increased to the limit value of 0.5 µg/ml, after ten consecutive exposures to ertapenem. ESBL-producing K. pneumoniae strain coded ERT-4 developed resistance to ertapenem on the sixth day of ertapenem exposure, while strain ERT-5 developed resistance to ertapenem on the fourth day of exposure. Isolate TZP-1 was susceptible to TZP and ertapenem after ten consecutive TZP exposures. The TZP-3 isolate had a TZP MIC of 128 ug/ml and an ertapenem MIC of 0.5 ug/ml at the end of the experiment. The RT-PCR was performed to all isolates after the completion of the experiments and bla OXA-48 gene was detected only in the ERT-2 isolate. No carbapenemase gene was detected in the other isolates. In summary; administration of TZP and meropenem upto 14 days did not cause fecal CRE colonization in mice, but TZP administration resulted with fecal ESBL-producing Enterobactarales colonization. In vitro experiments showed that carbepenem MIC was increased after carbapenem and TZP exposure, however, only bacteria exposed to carbapenem-resistant OXA-48-producing bacteria were found to develop resistance after repeated exposure to ertapenem. This study was supported by Hacettepe University Scientific Research Projects Coordination Unit on 15.12.2022. Project no: 20407