Standardization of C-Peptide Measurements

• Published in: Clinical chemistry (Baltimore, Md.) Md.), 2008-06, Vol.54 (6), p.1023-1026
• Main Authors: Little, Randie R, Rohlfing, Curt L, Tennill, Alethea L, Madsen, Richard W, Polonsky, Kenneth S, Myers, Gary L, Greenbaum, Carla J, Palmer, Jerry P, Rogatsky, Eduard, Stein, Daniel T
• Format: Article
• Language: English
• Subjects: Analytical, structural and metabolic biochemistry; Biological and medical sciences; C-Peptide - blood; Calibration; Chromatography, Liquid; Diabetes; Disease control; Fundamental and applied biological sciences. Psychology; Humans; Immunoassay; Immunoassay - methods; Investigative techniques, diagnostic techniques (general aspects); Kidney diseases; Laboratories; Least-Squares Analysis; Mass spectrometry; Mass Spectrometry - standards; Medical sciences; Methods; Peptides; Reference Standards; Standardization; Studies
• ISSN: 0009-9147; 1530-8561
• DOI: 10.1373/clinchem.2007.101287

C-peptide is a marker of insulin secretion in diabetic patients. We assessed within- and between-laboratory imprecision of C-peptide assays and determined whether serum calibrators with values assigned by mass spectrometry could be used to harmonize C-peptide results. We sent 40 different serum samples to 15 laboratories, which used 9 different routine C-peptide assay methods. We also sent matched plasma samples to another laboratory for C-peptide analysis with a reference mass spectrometry method. Each laboratory analyzed 8 of these samples in duplicate on each of 4 days to evaluate within- and between-day imprecision. The same 8 samples were also used to normalize the results for the remaining samples to the mass spectrometry reference method. Within- and between-run CVs ranged from 10% and from 18%, respectively. Normalizing the results with serum samples significantly improved the comparability among laboratories and methods. After normalization, the differences among laboratories in mean response were no longer statistically significant (P = 0.24), with least-squares means of 0.93-1.02. C-peptide results generated by different methods and laboratories do not always agree, especially at higher C-peptide concentrations. Within-laboratory imprecision also varied, with some methods giving much more consistent results than others. These data show that calibrating C-peptide measurement to a reference method can increase comparability between laboratories.