Hello, and thanks for joining me for this discussion about our research titled vancomycin and piperacillin-tazobactam, or CEPAPIM, and sepsis-associated AKI trajectory. Of no actual or potential conflicts of interest are disclosed today pertaining to this project or this presentation. To set the stage, we're all well aware that early and appropriate antimicrobial therapy has positive impacts on outcomes in patients with sepsis or septic shock. In our practice, many of us are probably accustomed to combination therapy with vancomycin and piperacillin-tazobactam, or VAN-TCP, or vancomycin and CEPAPIM, abbreviated VAN-FEP in this presentation, as choice empiric regimens in this population. While these are certainly appropriate broad coverage in most scenarios, we all know the growing yet controversial literature suggesting that an increased risk of acute kidney injury or AKI exists with a combination of VAN-TZP. Some studies have demonstrated a two- to three-fold increased risk with this combination compared to either agent alone or in combination VAN-FEP, while others find no appreciable differences between the incidence of AKI in critically ill patients receiving this regimen. Regardless of conflicting nature of these studies of VAN-TZP nephrotoxicity, there are some key points that present a significant practice gap. All current literature on this topic to date investigates the occurrence of incident AKI, so patients with preexisting AKI have been completely excluded. Interestingly, though, survey data suggests that the potential increased risk of incident AKI is extrapolated to patients with preexisting AKI as well. In a survey of those who believed that VAN-TZP posed an increased risk of nephrotoxicity, 64% recommended the combination less frequently in those with preexisting AKI, even though, as mentioned, these patients are excluded from all prior trials. A patient population particularly impacted by this practice are the 42% of sepsis patients presenting with AKI or sepsis-associated AKI, abbreviated SA-AKI in this presentation, who may receive alternative regimens due to concern for increased risk of AKI progression. While this may be a valid concern, there's no current data to suggest this risk actually exists. Therefore, our study aimed to evaluate the progression of kidney injury in patients with sepsis-associated AKI prescribed either VAN-TZP compared to VAN-FEP. This is a single-center retrospective cohort study conducted at a tertiary care academic medical center. These are medical intensive care unit patients admitted during the time period of January 1st, 2013 to July 31st, 2020. Inclusion criteria were adult patients 18 years of age or older who presented with sepsis or septic shock and concurrent AKI, again referred to as sepsis-associated or SA-AKI in this presentation. Patients also had to receive VAN-TZP or VAN-FEP. Patients were excluded if they were transferred from another hospital, died within 24 hours of ICU admission, or had a diagnosis of end-stage kidney disease prior to inclusion. Included patients were then divided into two groups based on the antimicrobial regimen they received within the first 72 hours. Patients had to have exclusively received either VAN-TZP or VAN-FEP for at least two consecutive days during this timeframe to be included in either the VAN-TZP or VAN-FEP groups respectively. We utilized the strict criteria to really try and clearly delineate the two groups and minimize overlapping beta-lactam exposure that could potentially lead to confounding in our results. The primary outcome for our study was maximum serum creatinine value during days two through seven, and secondary outcomes included several other kidney-related outcomes assessed at varying time points, as well as ICU mortality and ICU and hospital length of stay. For our analysis, we utilized Chi-square and Wilcoxon rink sum tests as appropriate. And due to the nature of the study, baseline serum creatinine was a key factor for determining the presence of AKI on admission. So for patients who did not have a baseline serum creatinine, we utilized multiple imputation with imputations based on the variables of age, race, sex, diabetes, and hypertension. To account for the potential impact of missing baseline serum creatinine, we also conducted a sensitivity analysis of the primary outcome, including only those patients with a documented baseline serum creatinine. Based on our power calculation, 176 patients were needed in both groups to detect a 20% difference in maximum serum creatinine value with 80% power. Here's a figure showing the application of inclusion and exclusion criteria in our study. So we started with 4,810 admissions that were initially screened, with 1,205 of those being septic patients meeting our preliminary inclusion criteria. Of those, 768 had acute kidney injury on admission and met the diagnosis of sepsis-associated AKI. After evaluating initial antimicrobial regimens, 480 patients were included in the cohort study, with 288 in the VAN-TZP group and 192 in the VAN-FEP group. This table represents some key demographics and baseline characteristics for the study cohort. However, you can find a more detailed and all-inclusive table using the QR code here on the slide below. Looking at the table, patients were well-matched between groups with no significant differences in any of the baseline characteristics. So we can see it was a relatively critically ill population, as evidenced by the higher SOFA score and a large proportion of patients requiring vasopressor support. A documented baseline serum creatinine was present for 48% of the population with similar rates of missing data in both groups. Looking at stage of AKI, stage 1 and stage 3 were the most common stages on admission, and there was no difference in the duration of antimicrobial exposure between the two groups. Our primary outcome of maximum serum creatinine during the first week was not different between the two groups. Similarly, no differences were seen in any of our secondary outcomes listed here on the slide either. In that sensitivity analysis that we conducted with patients only with documented baseline serum creatinine, a maximum serum creatinine value during the first week was 2.1 in 2 milligrams per deciliter, and progression of AKI was 23.5% versus 28.7% between the two groups. So again, no difference seen, very similar to our primary analysis. So in this study, we found no difference in kidney-related or clinical outcomes with the administration of VAN-TZP compared with VAN-FEP in patients with pre-existing AKI. These findings have important implications in that risk-benefit discussion regarding antimicrobial selection and empiric sepsis treatment. While VAN-FEP is the common alternative empiric regimen, utilizing this certainly introduces that cefepime-induced neurotoxicity concern, something that we already know patients with impaired renal function are at higher risk of. Additionally, VAN-FEP requires the addition of metronidazole for anaerobic coverage, leading to increased utilization of medication resources and additional antimicrobial exposure. Alternatively, to avoid the nephrotoxicity and neurotoxicity concerns altogether, vancomycin could be used in combination with miripenem. However, this is an extremely broad combination that likely would have severe antimicrobial resistance-related consequences if widely used. In line with antimicrobial resistance concerns, there's little data on the impact of exposure to short courses of multiple beta-lactams. So, for example, a patient is getting a couple of days of VAN-TZP, develops an AKI, and then a switch to VAN-FEP for definitive therapy. This multiple exposure pattern to beta-lactams could have unrecognized consequences that are ultimately being driven by this nephrotoxicity concern with VAN-TZP. Some strengths of our study include the novel population of patients with preexisting AKI, which again, as previously mentioned, have been excluded from all prior literature on this topic. We also set a relatively strict inclusion definition to clearly define specific beta-lactam exposure and limit potential crossover and confounding between the two groups. We also explored various kidney endpoints and at various time points to better evaluate the safety of this antimicrobial regimen in patients with AKI. Some limitations of our work include the single-center retrospective nature and the critically ill population, which may limit generalizability and application of these results to non-critically ill populations. We also didn't collect any kidney structural or functional biomarkers and solely relied on serum creatinine, which has well-identified limitations. Lastly, while we effectively qualified antimicrobial exposure, we did not quantify the number or doses or cumulative doses patients may have received or collect any therapeutic drug monitoring information, which may play a role in assessing any toxicity in this population. Moving forward, it would be interesting to conduct a similar study as ours, but incorporating kidney biomarkers in the assessment of acute kidney injury, especially considering a recent study looking at incident AKI, showed no difference in cis-statin C, a kidney function biomarker, despite an increase in serum creatinine and serum creatinine-defined AKI with a combination of Van-TZP compared to Van-FEP. Additionally, several mechanisms have been proposed to explain the sort of pseudonephrotoxicity that is hypothesized to occur with a combination of Van-TZP. However, it is unclear if similar mechanisms hold true in sepsis-associated AKI due to the unique pathophysiology, and this too would be an interesting area to focus on for future research. Before closing, I wanted to give a huge thank you to this research team, and especially Kaylee Whitenack, who's the first author on this publication, for bringing this great project to fruition. And again, thank you all for joining on this discussion today. Please feel free to connect with me or submit any questions you might have. Have a great day.

vancomycin

piperacillin-tazobactam

CEPAPIM

sepsis-associated acute kidney injury

antimicrobial selection