Evaluation of an in-clinic Serum Amyloid A (SAA) assay and assessment of the effects of storage on SAA samples

BACKGROUND An in-clinic assay for equine serum amyloid A (SAA) analysis, Equinostic EVA1, was evaluated for use in a clinical setting. Stability of SAA in serum samples was determined. METHODS Intra- and inter- assay variation of the in-clinic method was determined. The in-clinic method (EVA1) results were compared to a reference method (Eiken LZ SAA) with 62 patient samples. For samples with SAA concentrations within the assay range of EVA1 (10-270 mg/L), differences between the methods were evaluated in a difference plot. Linearity under dilution was evaluated in two samples. Stability of SAA in three serum pools stored at 4 degrees C and approximately 22 degrees C was evaluated with the reference method day 0, 1, 2, 4, 7, 17 and analysed with a two-way ANOVA. RESULTS The imprecision (coefficient of variation, CV) for the in-clinic method was acceptable at higher SAA concentrations with CV values of 7,3-12%, but poor at low SAA concentrations with CV values of 27% and 37% for intra- and inter-assay variation respectively. Recovery after dilution was 50-138%. The in-clinic assay and the reference method identified equally well horses with low (<10 mg/L) and high (>270 mg/L) SAA concentrations. Within the assay range of the in-clinic method, 10-270 mg/L, the difference between the two methods was slightly higher than could be explained by the inherent imprecision of the assays. There were no significant changes of serum SAA concentrations during storage. CONCLUSIONS The in-clinic assay identified horses with SAA concentrations of <10 mg/L and >270 mg/L in a similar way as the reference method, and provided an estimate of the SAA concentration in the range of 10-270 mg/L. The imprecision of the in-clinic method was acceptable at high SAA concentrations but not at low concentrations. Dilution of samples gave inconsistent results. SAA was stable both at room temperature and refrigerated, and thus samples may be stored before analysis with the reference method.