Cross-validation of commercial enzyme-linked immunosorbent assay and radioimmunoassay for porcine C-peptide concentration measurements in non-human primate serum.

Abstract

C-peptide concentration is widely used as a marker of insulin secretion and is especially relevant in evaluating islet graft function following transplantation, because its measurement is not confounded by the presence of exogenous insulin. To address the shortage of human islet donors, the use of porcine islets has been proposed as a possible solution and the stringent pig-to-non-human primate (NHP) model is often the most relevant for pre-clinical evaluation of the potential for diabetes reversal resulting from an islet xenograft. The Millipore radioimmunoassay (RIA) was exclusively used to measure porcine C-peptide (PCP) until 2013 when the assay was discontinued and subsequently a commercially available enzyme-linked immunosorbent assay (ELISA) from Mercodia has been widely adopted. Both assays have been used in pre-clinical trials evaluating the therapeutic potential of xenograft products in reversing diabetes in the pig-to-NHP model, to interpret data in a comparable way it may be useful to perform a harmonization of C-peptide measurements. We performed a method comparison by determining the PCP concentration in 620 serum samples collected from 20 diabetic cynomolgus macaques transplanted with adult porcine islets. All analyses were performed according to manufacturer instructions. With both assays, we demonstrated an acceptable detection limit, precision, and recovery. Linearity of the ELISA met acceptance criteria at all concentrations tested while linearity of the RIA only met acceptance criteria at five of the eight concentrations tested. The RIA had a detection limit of 0.16ng/mL, and recovery ranged from 82% to 96% and met linearity acceptance criteria at 0.35ng/mL and from 0.78 to 2.33ng/mL. The ELISA had a detection limit of 0.03ng/mL, and recovery ranged from 81% to 115% and met linearity acceptance criteria from 0.08 to 0.85ng/mL. Both assays had intra-assay precision <11% and inter-assay precision <14%. PCP concentration measured by ELISA demonstrated a significant correlation with RIA (R2 =.9721, P<.0001). This strong correlation supports use of the regression equation y=2.029x+0.0897 to transform ELISA data to RIA or inversely y=0.4930x-0.0456 to convert RIA data to ELISA for direct comparison between assays in the concentration range of 0-3.0ng/mL. Measured C-peptide concentration was lower with the ELISA than with the RIA; individual measurements plotted against the averages of the pair demonstrated that the variability from the mean strongly depended on increasing concentration. Porcine C-peptide can be reliably measured in NHP serum using the Mercodia ELISA, making this assay interchangeable with the Millipore RIA. Inherent differences in antibody affinity and calibration factors may explain the lower ELISA values as compared to the RIA; however without access to a traceable reference standard, it is not possible to determine which assay is most accurate. Regression modeling resulted in a correction factor appropriate for conversion of ELISA data to RIA-equivalent data facilitating comparison of assay results longitudinally and between groups.