RESUMO
Background: Prerenal acute kidney injury (prerenal AKI), hepatorenal syndrome (HRS-AKI), and acute tubular necrosis (ATN-AKI) are the various phenotypes of acute kidney injury, which are described in decompensated cirrhosis. It has therapeutic and prognostic implications. We aimed to evaluate the diagnostic utility of fractional excretion of sodium and urea (FENa and FEUrea) for differentiating AKI phenotypes. Methods: A prospective analysis was performed in 200 patients with decompensated cirrhosis with AKI to derive receiver operating curve, optimal cut-off, sensitivity, and specificity. These findings were validated in an independent cohort (n = 50) to differentiate ATN-AKI, HRS-AKI, and prerenal AKI. Results: The incidence of prerenal AKI, HRS-AKI, and ATN-AKI were 37.5%, 34%, and 28.5% in the derivation cohort and 28%, 38%, and 34% in the validation cohort respectively. The median FENa was significantly different in various phenotypes of AKI in the derivation and validation cohort (P 0.001); FEUrea was significantly different in the derivation cohort (P 0.0001), not in the validation cohort (P 0.106). The AUC for FENa (cut-off, sensitivity/specificity) was 86.6% (0.567, 89/71) and for FEUrea was 60.3% (34.73, 70/58) for ATN-AKI vs. non-ATN-AKI. The area under the curve for FENa to differentiate between HRS-AKI vs. non-HRS-AKI was 74.5%. FEUrea could not differentiate HRS-AKI vs. non-HRS-AKI (AUC 60.4%) satisfactorily. FENa and FEUrea were unable to differentiate between prerenal AKI and HRS-AKI (AUC <70%). Conclusion: Among cirrhotics FENa at admission is a simple, commonly available clinical tool that can be used to differentiate structural AKI from prerenal AKI and HRS-AKI. The newly derived lowered cut-off value of FENa makes the diagnosis of ATN-AKI easier, faster and thus obviates the need for extensive workup in a significant proportion of patients. FENa appears better than FEUrea in decompensated cirrhosis with AKI.
