Progression of albuminuria in patients with sickle cell anemia: a multicenter, longitudinal study

Abstract

Sickle cell nephropathy results in chronic kidney disease (CKD), which is associated with significant morbidity and mortality in sickle cell anemia (SCA). Albuminuria is an early manifestation of sickle nephropathy; however, little is known about progression of albuminuria or its correlation with glomerular filtration rate (GFR) decline or CKD. We studied nephropathy progression in 303 SCA participants in a prospective, multicenter, longitudinal study. We collected steady-state urine and serum samples yearly and assessed albumin/creatinine ratio (ACR), estimated GFR (eGFR), and SCA and nephropathy biomarkers. Participants with albuminuria (ACR ≥30 mg/g) for ≥2 annual measurements were classified as having persistent albuminuria (PA). At baseline (mean age, 21 years; range, 2-64 years), 32% had albuminuria. In longitudinal multivariate analysis, ACR was associated with sex, anemia, older age, and higher bilirubin and kidney injury molecule-1 levels. Albuminuria increased with age by 3.5 mg/g per year (P < .0001). Of 175 participants with ≥3 annual samples, 81% with baseline albuminuria ≥100 mg/g developed PA. Decreased eGFR and adult CKD were associated with PA (P = .002 and P = .02, respectively), but not with baseline albuminuria. Rate of eGFR decline was steeper among adults (but not children) with albuminuria, compared with those without (P = .02). Participants with PA were more likely to have rapid eGFR decline compared with those without (P = .03). In this longitudinal study, albuminuria progressed with age, and adults with albuminuria had worse eGFR decline than those without. Albuminuria ≥100 mg/g predicted PA, which was associated with rapid eGFR decline and CKD development in adults with SCA. This trial was registered at www.clinicaltrials.gov as #NCT02239016.

Graphical abstract

Figure 1.

Longitudinal association of ACR with age. ACR progression with age including all participants in the study. Center line is a fitted regression line, with 95% confidence interval [CI] upper and lower bound lines. Each filled circle represents a single time point. Darker-appearing circles are the result of overlaying time points.

Figure 2.

Baseline ACR determines the probability of persistent albuminuria. The probability of PA based on baseline ACR.

Figure 3.

Progression of eGFR-CysC. The rate of eGFR-CysC decline in children (A) and adults (B) based on presence or absence of albuminuria. Red lines represent participants with albuminuria, and blue lines represent participants without albuminuria. Filled circles represent mean eGFR value spanned by standard error of the mean. The median yearly eGFR decline (mL/min/1.73 m² per year) is indicated for both groups. Association between change in eGFR-CysC and ACR as a continuous variable among adult participants with PA (C) and non-PA (D). Regression lines with 95% CIs are shown in red for adults with PA and black for adults with non-PA. (E) Linear regression of eGFR-CysC vs age in children (<18 years). (F) Linear regression of eGFR-CysC vs age in adults.