#1708 Mortality and kidney function recovery in patients with sepsis-associated vs non-septic acute kidney injury treated with CRRT

Abstract Background and Aims It was recently suggested that sepsis-associated acute kidney injury (SA-AKI) is characterized by different pathogenesis and outcomes compared to non-Septic AKI (NS-AKI). However, limited data are available focusing on the clinical outcomes of SA-AKI versus NS-AKI. The aim of this study is to examine risk factors and outcomes of SA-AKI vs NS-AKI among critically ill patients who developed AKI requiring renal replacement therapy (RRT). Methods We performed a single-center retrospective analysis, examining all patients admitted to ICU who developed AKI requiring RRT at Policlinico of Bari, from January 2021 to July 2023. We classified patients in two groups based on the leading cause of AKI: NS-AKI and SA-AKI. We collected clinical and laboratory data, including maximum AKI stage reached during hospitalization, RRT indications, modality and duration. The primary outcome was to assess mortality rate between SA-AKI and NS-AKI groups. The cumulative 28-day survival probability was calculated using the Kaplan–Meier method. The analysis of the risk factors associated with in-hospital mortality was performed using Cox's proportional hazard regression models. Kidney Functional Recovery (KFR) from, RRT discontinuation and length of ICU and hospital stay were also analyzed. Results 320 patients were included in the study; 131 patients developed SA-AKI (40.9%), while the remaining 189 were classified as NS-AKI (59.1%). Median age was 67 years (IQR 57-73) with a predominance of male gender (67.8%), without significant differences between groups. A significant percentage of patients had an history of cardiovascular disease in NS-AKI group compared to SA-AKI group (54.5% vs 29%, p 0.001). Conversely, the proportion of patients with a history of chronic respiratory disease was significantly higher in the SA-AKI group (23.7% vs 9.5%, p = 0.001). 268 patients (83.8%) developed AKI Stage 3 and 40 patients AKI stage 2 (12.5%) at RRT initiation. All patients were treated with continuous modalities (CRRT). There was no significant difference in CRRT duration between groups (median days 5, IQR 2-12) (Table 1). The timing of CRRT initiation from ICU admission was significant different and we reported a late CRRT initiation in the SA-AKI group (3 vs 1 days, p = 0.002). The median length of ICU stay was 12 days (IQR 4-25) and did not differ between groups. 184 patients (57.5%) died at 28 days, with a higher percentage in the SA-AKI (70.2 vs 48.6%, p 0.001). Similarly, in-hospital mortality was reported in 221 patients (69%), with impressive percentages in the SA-AKI group (82.45 vs 69.3%, p 0.001) (Fig. 1). KFR at 28 days from CRRT initiation was significantly higher in the NS-AKI (28.5%) compared to the SA-AKI group (13.7%, p = 0.002). Median time of KFR was similar between groups (p = 0.833). Multivariate Cox Regression analysis showed that only SA-AKI (HR 1.442, 95% CI 1.052-1.978, p = 0.023) was independently associated with increased risk of mortality. Conclusions Sepsis-associated AKI was associated with high mortality rate and lower likelihood of renal recovery compared to non-septic AKI in critically ill patients requiring RRT. These findings emphasize the importance of identifying specific AKI phenotypes characterized by different pathophysiology and clinical outcomes.