Early prediction of COVID-19-associated acute kidney injury: Are serum NGAL and serum Cystatin C levels better than serum creatinine?

Background: Coronavirus disease-2019 (COVID-19) is associated with a high risk of acute kidney injury (AKI), often requiring renal replacement therapy (RRT). Serum Cystatin C (sCysC) and serum Neutrophil Gelatinase-Associated Lipocalin (sNGAL) are emerging biomarkers for kidney injury, and were suggested to be superior to serum creatinine (sCr) in several clinical settings. Moreover, elevated sCysC is associated with disease severity and mortality in COVID-19. We aimed to assess the utility of sCysC and sNGAL for predicting COVID-19-associated AKI, need for RRT, and need for intensive care unit (ICU) admission, when measured at presentation to the emergency department (ED). Methods: Patients presenting to the ED with laboratory-confirmed COVID-19 were included. The primary outcome was development of COVID-19-associated AKI, while the secondary outcomes were need for RRT and ICU admission. Results: Among 52 COVID-19 patients, 22 (42.3%) developed AKI with 8/22 (36.4%) requiring RRT. Both sCr and sCysC demonstrated excellent performance for predicting AKI (AUC, 0.86 and 0.87, respectively) and need for RRT (AUC, 0.94 and 0.95, respectively). sNGAL displayed acceptable performance for predicting AKI (AUC, 0.81) and need for RRT (AUC, 0.87). Conclusions: SCr and sCysC measured at ED presentation are both highly accurate predictors of AKI and need for RRT, whereas sNGAL demonstrated adequate diagnostic performance. While sCyC was previously shown to be superior to sCr as a diagnostic biomarker of kidney injury in certain etiologies, our findings demonstrate that sCr is comparable to sCyC in the context of predicting COVID-19-associated AKI. Given the high sensitivity of these biomarkers for predicting the need for RRT, and as sCysC is associated with mortality in COVID-19 patients, we recommend their measurement for enabling risk stratification and early intervention.

Automated summary

This study aimed to evaluate the utility of sCysC and sNGAL measured at the emergency department (ED) presentation for predicting COVID-19-associated AKI, need for RRT, and ICU admission. The results showed that both sCr and sCysC had excellent performance in predicting AKI and need for RRT, while sNGAL had acceptable diagnostic performance.