How Accurate is Your Sclerostin Measurement? Comparison Between Three Commercially Available Sclerostin ELISA Kits

Abstract

Sclerostin, bone formation antagonist is in the spotlight as a potential biomarker for diseases presenting with associated bone disorders such as chronic kidney disease (CDK-MBD). Accurate measurement of sclerostin is therefore important. Several immunoassays are available to measure sclerostin in serum and plasma. We compared the performance of three commercial ELISA kits. We measured sclerostin concentrations in serum and EDTA plasma obtained from healthy young (18-26 years) human subjects using kits from Biomedica, TECOmedical and from R&D Systems. The circulating sclerostin concentrations were systematically higher when measured with the Biomedica assay (serum: 35.5 ± 1.1 pmol/L; EDTA: 39.4 ± 2.0 pmol/L; mean ± SD) as compared with TECOmedical (serum: 21.8 ± 0.7 pmol/L; EDTA: 27.2 ± 1.3 pmol/L) and R&D Systems (serum: 7.6 ± 0.3 pmol/L; EDTA: 30.9 ± 1.5 pmol/L). We found a good correlation between the assay for EDTA plasma (r > 0.6; p < 0.001) while in serum, only measurements obtained using TECOmedical and R&D Systems assays correlated significantly (r = 0.78; p < 0.001). There was no correlation between matrices results when using the Biomedica kit (r = 0.20). The variability in values generated from Biomedica, R&D Systems and TECOmedical assays raises questions regarding the accuracy and specificity of the assays. Direct comparison of studies using different kits is not possible and great care should be given to measurement of sclerostin, with traceability of reagents. Standardization with appropriate material is required before different sclerostin assays can be introduced in clinical practice.

Keywords: Clinical utility; ELISA; Metabolic bone disease; Sclerostin.

Fig. 1

Passing-Bablock regression analysis for serum (left panel) and EDTA (right panel) samples comparing the three different ELISA kits for circulating sclerostin measurements. Dash line represents the fitted regression line; dark grey dotted lines represent upper and lower 95 % confidence and light grey dotted line represent the identity line

Fig. 2

Bland–Altman plots for sclerostin concentrations in EDTA plasma comparing the three different ELISA kits. R&D systems showed little bias when compared to TECOmedical while both R&D Systems and TECOmedical assays showed a negative bias (and wide CI) compared to Biomedica; bias present mainly at the highest concentrations of sclerostin

Fig. 3

Bland–Altman plots for sclerostin concentrations in serum comparing the three different ELISA kits. R&D systems showed a negative bias when compared to TECOmedical as well as Biomedica that proportionally increased with increasing concentrations of sclerostin. TECOmedical showed a negative bias compared to Biomedica which affected mainly the highest concentrations of sclerostin

Fig. 4

Bland–Altman plots comparing sclerostin concentration in serum versus EDTA plasma using the three different ELISA kits. TECOmedical showed very little bias between serum and EDTA samples. There was a systematic and proportional negative bias with the R&D Systems assay (from −11 to −38 pmol/L). The bias was present mainly for the high concentrations of sclerostin using the Biomedica assays