Purpose To assess the prognostic ability of CysC for early detection of AKI and to examine its incremental value over traditional markers for both AKI and all-cause mortality.
Methods Cohort analysis of 169 patients (81.1% male, mean±SD age=70.18±13.06 years old) admitted for AHF with ejection fraction (EF) <35% (Mean±SD=30.40±8.45%) in a single centre coronary ICU and who had available measurements of admission’s blood CysC, Creat and urea. AKI and all-cause mortality (in-hospital and long-term, 25.36±26.26 months) were analyzed. We determined AKI by the KDIGO (Kidney Disease: Improving Global Outcomes) definition. Differences between groups were analyzed with a Student’s t-test. Difference in the incidence of AKI in patients with and without chronic kidney disease (CKD) was analyzed with a chi-square test. Receiver operating characteristic (ROC) curves were used to determine the discriminator power of these measures regarding AKI and all-cause mortality. Area under the curve (AUC) of the different ROC curves were compared to establish superiority of one test (DeLong method). Statistical significance was accepted for p values <0.05.
Results An acute decompensation of heart failure was the most common presentation (74.6%), 41.4% had CKD and mean±SD length of stay was 12.44±9.89 days. In-hospital AKI occurred in 49.70% patients and in-hospital and long-term death in 10.10% and 46.20% patients, respectively. CKD patients were more likely to develop AKI (60% vs 42.4%, p=0.024). Mean levels of CysC were significantly higher in were higher in CKD patients (2.33±1.14mg/L vs 1.35±0.58mg/L, p<0.000) and in patients who developed AKI (1.99±1.04mg/L vs 1.53±0,86mg/L, p=0,002) but no differences were observed between survivors and non-survivors. CysC significantly improved AUC for the prediction of AKI (AUC=0.668, 95%CI=0.586–0.750, p=0,000) when comparing to urea’s and Creat’s AUC. When comparing patients with and without CKD, CysC is a significant predictor of AKI only in those with preserved renal function (AUC=0.647, p=0.013 vs AUC=0.599, p=0.163). For both in-hospital and long-term mortality urea had a greater AUC (for in-hospital mortality AUC=0.677, 95%CI=0.586-0.816, p=0.017; for long-term mortality AUC=0.722, 95%CI=0.639-0.805, p=0,000) with significant superiority comparing to CysC and Creat (DeLong method).
Conclusion In our sample, admission CysC was a better predictor of in-hospital AKI and it added prognostic value particularly in patients with preserved renal function. It may be used as an early surrogate of AKI and promote kidney protective measures in such patients.