Background: Soluble suppression of tumorigenicity (sST2) constitutes a novel biomarker with diagnostic and prognostic implications in several diseases. However, recent evidence suggests that different enzyme-linked immunosorbent assay (ELISA) kits could result in diverging serum concentrations measured.
Methods: Serum concentrations of sST2 were measured in blood of 215 patients with aortic valve stenosis using two commercially available ELISA-assays (Presage® ST2 assay and R&D). Passing and Bablok regression analysis, Bland-Altman plot, and correlation analysis were conducted.
Results: Values obtained by Presage® were 1.9-fold higher than concentrations measured by R&D, with a mean bias of 14,489 pg/mL between both assays. The most extreme deviations were observed in values below the median of concentrations measured by the R&D assay (21.4%, p < 0.0001).
Conclusions: Our findings suggest a constant difference and a proportional bias between both investigated assays could be of special importance in circumstances where cutoffs with prognostic relevance have been calculated previously. In order to interpret sST2 concentrations correctly, the clinician should be aware of these deviations between different ELISA kits.