Validation of an automated immune turbidimetric assay for serum gelsolin and its possible clinical utility in sepsis

BackgroundStudies showing the potential predictive value of the actin-binding protein gelsolin, in critically ill patients are scarce. Moreover, even up to now a rapid automated measurement of gelsolin has still remained a challenge. Therefore, we developed and validated an automated serum gelsolin immune turbidimetric assay for possible clinical use. MethodsValidation of serum gelsolin assay was performed on a Cobas 8000/c502 analyzer (Roche) according to the second edition of Eurachem guidelines. Furthermore, we also studied the diagnostic value of serum gelsolin in sepsis when investigating sera of septic (n=25), systemic inflammatory response syndrome (SIRS; n=8) and control patients (n=14). We compared our previously published Western blot data with those of the new turbidimetric assay. ResultsThe sample volume was 7L and the assay time was 10minutes. The detection limit was 0.72mg/L, intra- and inter-assay imprecision remained in most cases less than 5% expressed as CV. Recovery was found to be 84.56%-93.52% and linearity study gave an appropriate correlation coefficient by linear regression analysis (r(2)=.998). Septic patients exhibited lower (P=.015) first-day serum gelsolin levels than SIRS patients, which confirmed our previous Western blot results. The determined cut-off point for serum gelsolin was 14.05mg/L (sensitivity: 75%; specificity: 60%) when investigating its diagnostic value in sepsis. ConclusionBased on the results, our immune turbidimetric measurement offers a rapid and accurate quantitation of gelsolin in human serum samples. Serum gelsolin seems a promising additional diagnostic marker of sepsis which has to be further investigated.