RESUMO
Canine serum amyloid A (SAA) is a useful diagnostic marker of systemic inflammation. A latex agglutination turbidimetric immunoassay (LAT) was validated for automated measurements. The aim of the study was to evaluate the clinical applicability of SAA measured by the LAT. SAA was measured in 7 groups of dogs with and without systemic inflammation (n=247). Overlap performance was investigated. Diagnostic performance was compared to body temperature and leukocyte markers. Clinical decision limits for SAA were estimated. In dogs with neurological, neoplastic or gastrointestinal disorders (n=143), it was investigated whether a higher proportion of SAA positive dogs could be detected in cases of complications with risk of systemic inflammation. Significantly higher concentrations of SAA were measured in dogs with (range [48.75; 5,032 mg/l]), compared to dogs without systemic inflammation [0; 56.4 mg/l]. SAA was a more sensitive and specific marker of systemic inflammation (area under the receiver-operating characteristic curve (AUC) 1.00), compared to body temperature (0.6) and segmented neutrophils (best performing leukocyte marker, 0.84). A clinical decision limit of 56.4 mg/l was established giving close to perfect discrimination between dogs with and without systemic inflammation. Higher proportions of SAA-positive dogs were observed in dogs with neurological, neoplastic and gastrointestinal disorders with complications known to increase risk of systemic inflammation, compared to uncomplicated cases. The automated LAT makes SAA applicable as a relevant diagnostic marker of systemic inflammation in dogs for routine random-access real-time use in a general clinical setting.