B-347 Performance Validation of an In-house Assay for Soluble Mesothelin Related Peptides

Abstract

Abstract Background Mesothelin is a glycosylated phosphatidylinositol glycoprotein located on the surface of mesothelial cells. Mesothelin is released from the cell in a form denoted as Soluble Mesothelin Related Peptides (SMRP), that can be measured in the blood. Although mesothelin is produced by normal cells, studies have shown an overexpression in individuals suffering malignant mesothelioma, with an increase in serum SMRP concentration associated with the disease. The in-house ARUP MESOMARK® ELISA is a 96-well microtiter plate formatted sandwich immunoassay for the quantitative measurement of SMRP in human serum. The ELISA incorporates two separate monoclonal antibodies specific for SMRP (including megakaryocyte potentiating factor); one (4H3) bound to the microtiter well for capturing SMRP and the second (OV569), enzyme labeled with horseradish peroxidase (HRP), for bound SMRP detection. After several incubation and wash steps to remove unbound materials, color is produced by HRP turnover of a chromogenic substrate during a final incubation step. A direct relationship between the color intensity and the SMRP concentration of the specimen is generated, with results expressed as nmol/L SMRP. Methods Deidentified residual serum specimens sent to ARUP Laboratories for routine testing, as well as serum specimens from healthy volunteers, were collected under University of Utah’s Institutional Review Board approved protocols. Commercial kit (Fujirebio Diagnostics, Inc., Malvern, PA) testing was completed according to manufacturer’s protocol. ARUP ELISA measurements were conducted according to an established in-house protocol. ARUP ELISA performance characteristics evaluated included a method comparison against the commercial assay, linearity, recovery, precision, analytical sensitivity, and dilution studies. A reference limit was also verified. Results A method comparison study against the commercial SMRP ELISA generated a slope of 0.989, intercept of 0.0422, r2 of 0.998 and bias of -1.2% (n = 45, Deming regression, Bland-Altman analysis). Clinical agreement near the 1.5 nmol/L cutoff was 100%. The limit of detection was 0.3 nmol/L (20 determinations each of a blank and low-level serum pool). Linearity was established by combining serum specimens with high and low SMRP concentrations at different ratios, creating 10 specimens of varying SMRP concentrations. Each specimen was tested in triplicate. Regression analysis (measured vs expected concentration) produced a slope of 1.002, intercept of -0.739 and r2 of 0.998, with percent recoveries ranging 86.1 to 100.0%. Precision was determined from two serum pools of differing SMRP concentrations tested over 5 days, four replicates per pool per day. Repeatability and within-laboratory CVs were 3.0 and 5.0% at 1.6 nmol/L, and 1.7 and 1.7% at 17.3 nmol/L, respectively. Back-calculated results from a 10-fold dilution study of four serum specimens were within -9.8 to -8.1% of their corresponding neat values. The previously established reference limit of 1.5 nmol/L was verified for the ARUP ELISA, with 95% of results below the cutoff value (n = 20, sera from healthy volunteers). Conclusions The in-house ARUP MESOMARK ELISA demonstrates acceptable performance for quantifying SMRP in human serum and compares favorably to a well-established commercial ELISA. A reference limit of 1.5 nmol/L was also verified. Overall, the ARUP MESOMARK ELISA is acceptable for serum SMRP testing.