Renal perfusion status remains poorly studied at the bedside during sepsis associated acute kidney injury (AKI). The aim of the study is to examine renal cortical and medullary perfusion using renal contrast enhanced ultrasound (CEUS) in septic patients. In this single-center, prospective longitudinal study, septic patients were enrolled. Renal ultrasonography was performed within 24 hours of ICU admission (D1), then repeated at D3, D5 and D7. Each measurement consisted of three destruction replenishment sequences that were recorded for delayed analysis with dedicated software (Vuebox). Renal cortex and medulla perfusion were quantified by measuring time to peak (TTP). Receiver operating characteristic (ROC) analysis was used to evaluate 28-day renal prognosis. The study included 149 septic patients, including 70 non-AKI patients and 79 AKI patients. Both renal cortical and medullary TTP was longer in the AKI group than in the non-AKI group. The difference of TTP between renal cortex and medulla in AKI group was higher than that in the non-AKI group (p = 0.000). Medullary TTP on day 3 had the best performance in predicting the prognosis of 28-day renal function (AUC 0.673, 95% confidence interval 0.528-0.818, p = 0.024), and its cut-off value was 45 s with a sensitivity 52.2% and a specificity of 82.1%. Cortical TTP on day 3 also had the performance in predicting the prognosis of 28-day renal function (AUC 0.657, 95% confidence interval 0.514-0.800, p = 0.039), and its cut-off value was 33 s with a sensitivity 78.3% and a specificity of 55.0%. Renal medullary perfusion alterations differ from those in cortex, with the medulla is worse. Simultaneous and dynamic assessment of cortical and medullary microcirculatory flow alterations necessary. TTP on day 3, especially medullary TTP, seems to be a relatively stable and useful indicator, which correlates with 28-day renal function prognosis in septic patients. Early correction of renal cortical and medullary perfusion alterations reduces the incidence of adverse renal events.
Read full abstract