Comparison of serum protein electrophoresis by agarose gel and capillary zone electrophoresis in a clinical setting

Publisher Website
Google Scholar
Abstract
Capillary zone electrophoresis (CZE) offers the potential for automating serum protein electrophoretic analysis traditionally performed on standard thin‐layer agarose gels. The following describes the use of CZE compared to agarose gel electrophoresis (AGE) for the detection of dysproteinemia and paraproteinemia in a clinical study involving 240 patients. The study includes within‐run and between‐run reproducibility data on the Paragon CZE® 2000 Clinical Capillary Electrophoresis System, in addition to concordance data between the two methodologies. Paraprotein quantitation studies comparing AGE versus CZE were also performed. Reproducibility for the automated CZE system was superior to the AGE system. Improved reproducibility for the CZE method is largely due to measuring protein absorbance directly at 214 nm versus the traditional AGE method that measures the amount of dye adsorbed to protein. Reproducibility data as percent coefficient of variance (% CV) for the five classic bands in a normal control serum for between‐run precision ranged from 1.2 to 4.5% for CZE compared to AGE, which ranged from 3.8 to 8.0% CV. Concordance studies between AGE and CZE involving dysproteinemias including hypogammaglobulinemia, polyclonal and monoclonal gammopathies, acute and chronic inflammation, nephrosis, hepatodegenerative disease, cirrhosis, and iron deficiency anemia showed 96% agreement. Paraprotein classification, which compared the CZE immunosubtraction method to immunofixation electrophoresis (IFE) on agarose, showed 100% agreement. Certain dysproteinemias involving beta lipoprotein were in partial concordance due to the inability of the CZE procedure to detect this component. Detection limits for monoclonal gammopathies, providing they were not comigrating with other proteins, were IgG 50 mg/dL, IgM 75 mg/dL, and IgA 75 mg/dL. Paraprotein quantitative studies between the two methods showed less than a ± 0.2 g/dL variation.