B-202 Improving Performance of an LC-MS C-peptide Reference Method

Abstract Background C-peptide is a useful biomarker for distinguishing among different types of diabetes and is a proxy measure of endogenous insulin production, useful in diabetes management. Additionally, C-peptide tests can be used to diagnose non-diabetic hypoglycemia and insulinoma. However, differences among clinical assays can limit the application of C-peptide testing. To address this issue, efforts are underway to standardize manufacturer assays with serum-based secondary reference materials where values are assigned by a reference LC-MS method. Clinical evidence suggests that patients with diabetes can benefit from residual beta-cell function, yet many clinical assays have high variability at lower levels of C-peptide. Here we explore different approaches for improving sensitivity and selectivity of C-peptide reference measurements with a focus at the lower end of the measurement range. Methods Two methods were developed to evaluate and determine the most effective approach. The first method involved spiking serum samples with isotope-labeled c-peptide (Sigma) and then enriching and enzymatically digesting them. The enrichment process included removing abundant serum proteins by using methanol, followed by centrifugation and filtration through C18 cartridges. The resulting solution was then passed through a strong anion-exchange column after correcting its pH with ammonium formate buffer. The eluent was evaporated, and the residue was dissolved in ammonium carbonate buffer. Enzymatic digestion was performed by adding dithiothreitol and iodoacetamide followed by addition of Glu-C. The digested peptides were then analyzed by LC-MS/MS (QTRAP 6500+) in MRM mode. The second method followed the same steps as the first method, except that the evaporation and digestion step were skipped, and MRM was performed on the intact C-peptide chain. Two separate calibration curves were constructed for each method using isotope labeled c-peptide (Sigma) and certified reference material (National Metrology Institute of Japan) as native c-peptide. Results The first method, which measures intact C-peptide using MRM, had a slope of 1.14 and an intercept of −0.01 nmol/L (Deming regression, R = 0.917) when compared to the reference method. The method demonstrated CV within 0.4%–11.9%. The second method, which quantifies digested peptides, showed a slope of 1.02 and an intercept of −0.21 nmol/L compared to the reference method (Deming regression, R = 0.995). The method demonstrated CV within 4.4–20.0%. Controlling the three MRM transitions improved the specificity and accuracy of the results. For the first method two MRM transitions were adequate. Both methods showed no significant matrix effects. The first method is simpler to implement due to a smaller number of separation steps. Conclusion We have explored two new alternative workflows for C-peptide quantitation that include first, an enzymatic digestion then MRM quantitation and second, MRM quantitation of intact C-peptide. To improve the analytical performance of the methods, further optimization is still required.