Serum Cystatin-C Levels in Infants with Sickle Cell Anemia: Baseline Data from the BABY HUG Trial

Background: BABY HUG is an NHLBI/NICHD sponsored double-blind placebo-controlled trial (NCT00006400) testing the hypothesis that hydroxyurea therapy (HU), if started early in life, will prevent or postpone organ damage to the spleen and kidney in infants with sickle cell anemia (SCA). All BABY HUG subjects at entry have GFR measured by ^99mTc-DTPA radioisotope clearance (DTPA) and estimated using the Schwartz formula, a calculation based on serum creatinine (Cr) and height of the patient (GFR = κ (height)/Cr); κ=0.55 for children age >1 to 13 yr or 0.45 for term infants to age 1 yr. Measurement of GFR in infants is problematic due to difficulties in obtaining urine specimens and venous access. While Cr is a well-established marker for GFR it is not independent of body mass and may be secreted by renal tubules, leading to an overestimate of GFR. Cystatin C (CysC) is a cysteine protease inhibitor produced by all human nucleated cells and freely filtered by the kidney. Serum levels are not affected by muscle mass or gender and stable from age 18 mo to 50 yr, with higher levels in newborns reflecting a lower GFR. CysC offers a more practical and perhaps more accurate measure of GFR than Cr, especially in children with SCA. Normal non-SCA values are 0.5 to 1.4 mg/L. Formulas are published to translate serum levels of CysC to GFR as conventionally reported. Our aim was to determine usual and mean CysC levels in infants with SCA and compare CysC to DTPA and Cr-based assessments of GFR.

Methods: Sera obtained and frozen during the eligibility screening phase of BABY HUG were used to determine CysC levels by particle enhanced immune nephelometry.

Results: A total of 152 sera from infants age 9–17 mo (mean 13.5) were available. CysC levels ranged from 0.532 to 1.369 mg/L (mean 0.92; median 0.907), approximating the normal range. CysC was strongly and inversely associated with GFR estimated by DTPA (R²=0.086; p=0.001). CysC was also significantly associated with age (inversely, p=0.02), Schwartz GFR (using κ=0.55) (inversely, p=0.007) and Cr (p=0.01), but not with Hb, HbF or WBC count. GFR determined using two CysC-based formulae and the Schwartz formula were compared to GFR by DTPA.

GFR determined using CysC-based formula 2: GFR = antilog {1.92 + [1.123 × log (1/CysC)]} gave mean values closest to DTPA and similar to published estimates of GFR in non-SCA infants. The Schwartz formula overestimated GFR, especially when the higher κ was used. While correlations are low due to high GFR variability, the four GFR determinations all have a significant association with DTPA. However, because the slopes of the regression lines are not close to 1 (line of equality), measurement agreement with DTPA is poor.

Conclusion: The Schwartz formula overestimated GFR in BABY HUG subjects. CysC-based GFR formulae underestimate GFR compared to concurrent DTPA values, but results are similar to published norms. Reports of hyperfiltration, based on creatinine-based determination of GFR, may be exaggerated. The impact of hydroxyurea therapy on CysC levels, if any, will be apparent at completion of the trial in late 2009.