Predictive Capacity of Plasma and Urinary Biomarkers for Persistent Albuminuria in Patients with Sickle Cell Disease

Introduction: Chronic kidney disease (CKD) is common in sickle cell disease (SCD) and is associated with increased mortality and morbidity. As albuminuria and changes in estimated glomerular filtration rate (eGFR) occur well after substantial structural and functional tissue damage have occurred, novel biomarkers are needed to predict CKD prior to its development. In this multicenter study, we investigated the association of multiple plasma and urine biomarkers with persistent albuminuria (PA) and their potential in predicting PA in patients with SCD.

Methods: Adult subjects with HbSS or HbSb ⁰ thalassemia were enrolled in a prospective, longitudinal cohort study of CKD in SCD. We excluded patients with diabetic nephropathy, cancer, connective tissue disease, other glomerular diseases, end-stage kidney disease on dialysis, known infection with hepatitis (B or C) or HIV, and who had undergone bone marrow transplantation and solid organ transplantation. Participants were enrolled during routine clinical visits in the non-crisis, “steady states.” PA was defined as at least 2 of 3 spot urine albumin-creatinine ratio (ACR) values ≥30 mg/g or a single value ≥100 mg/g. At baseline evaluation, plasma was analyzed for biomarkers of coagulation activation (D-dimer, thrombin antithrombin complex [TAT], ratio of cleaved high molecular weight kininogen-to-intact high molecular weight kininogen [cHK/iHK]), endothelial dysfunction (endothelin-1 [ET1], vascular endothelial growth factor [VEGF], vascular cell adhesion molecule-1 [VCAM-1]), and urine was analyzed for kidney injury molecule 1 (KIM-1), nephrin and angiotensinogen (AGT) using commercially available ELISA. All urinary biomarkers were normalized to urine creatinine. For baseline values, a standard t-test was performed for continuous variables and a chi-square test was performed for categorical variables. Logistic regression analysis was performed to assess univariate predictors of research biomarkers for PA.

Results: Of 278 evaluated subjects, 104 (37.4 %) had PA with a median ACR of 194.5 mg/g [IQR: 102.4, 487.1], and 174 subjects (62.6 %) did not have PA (non-PA) with a median ACR of 13.0 mg/g [IQR: 7.4, 22.4]. Subjects with PA were older, had lower hemoglobin and fetal hemoglobin, but higher systolic and diastolic blood pressures than those without PA (Table 1). Plasma VCAM-1, urinary KIM-1, and urinary AGT were significantly associated with PA with odds ratios of 1.001 (AUC: 0.636; 95% CI: 1.00, 1.002), 1.737 (AUC: 0.661; 95% CI: 1.226, 2.575), and 1.053 (AUC: 0.71; 95% CI: 1.029, 1.084), respectively (Table 2). Surprisingly, ET1 and urinary nephrin-creatinine ratio were not significantly associated with PA in this cohort.

Conclusion: Urinary AGT exhibited the highest sensitivity for PA, while urinary KIM-1 was the most specific. The relatively modest sensitivity of KIM-1, a marker of renal tubular dysfunction, is likely because albuminuria in SCD results from both glomerular and tubular dysfunction. These results highlight the potential importance of urinary AGT as a key biomarker in identifying individuals at risk for PA. Our ongoing studies will evaluate the predictive capacity of these biomarkers for development of PA in SCD.