QUANTITATION OF INSULIN BY CAPILLARY ELECTROPHORESIS AND HIGH-PERFORMANCE LIQUID-CHROMATOGRAPHY - METHOD COMPARISON AND VALIDATION

Two validated high-performance liquid chromatography (HPLC) and capill ary electrophoresis (CE) assays for insulin are compared. Both methods are selective and robust, with best reproducibility in a 0.9% sodium chloride solution, pH 7-8, as the sample solvent, It is shown that, be sides sample solvent and buffer pH, acetonitrile volatility is a cruci al point for reproducibility. Its high importance for selectivity and ruggedness is also stressed. Trends have been extensively investigated and characterized. The separation efficiency is better for the CE met hod. Furthermore, the analysis time of the CE method is up to four tim es shorter than the respective parameter in HPLC and the acetonitrile consumption is more than 100-fold less, Earlier works stated that all relevant precision data, such as relative signal standard deviation (1 .6% R.S.D. for peak areas, n = 20), precision of the analytical result and the limit of quantitation, were about a factor of two worse than for corresponding HPLC data (0.8% R.S.D. for peak areas, n = 19). This CE precision was further improved using relative instead of absolute peak areas, which compensate for the injection error (1.3% R.S.D. for relative peak areas, n = 20). If acetonitrile evaporation is avoided, by covering the buffer with mineral oil, reproducibility is even bette r than with the HPLC assay (0.5% R.S.D. for relative peak areas, n = 6 0). (C) 1997 Elsevier Science B.V.