Comparison of Two Ion-Exchange HPLC Methods for HbA1c Quantitation

This study characterized the analytical performance of the Tosoh Automated Glycohemoglobin Analyzer HLC-723GR01 (GR01), the newest-generation hemoglobin A1c (HbA1c) analyzer from Tosoh, relative to the predicate Tosoh HLC-723G8 glycohemoglobin analyzer (G8).

Performance of the GR01 was assessed through precision, linearity, method comparison, and variant interference studies in whole blood and hemolysate. The G8 was the comparator method for all studies. Variant interference studies included specimens with heterozygous hemoglobin variants C, D, E, and S, as well as elevations of hemoglobin A2. Interference studies for elevated hemoglobin F were also conducted using cord blood.

Total imprecision for the GR01 at HbA1c concentrations ranging from 5.0% to 12.0% (31-108 mmol/mol) was approximately 1% CV, meeting common acceptance limits applied to HbA1c clinical testing and method certification. Linearity studies also met acceptance criteria, with an average of 0.1% bias from the expected HbA1c concentration across the analytical measurement range. All paired GR01 and G8 HbA1c measurements in nonvariant specimens showed relative differences less than 5%; mean differences (in %HbA1c units) were 0.03% in whole blood (SD 0.07%) and 0.02% HbA1c in hemolysate (SD 0.07%). Across all variants, the mean difference between GR01 and G8 measurements was 0.1% (SD 0.25%).

The GR01 showed excellent correlation with the predicate G8, including in specimens containing high concentrations of common heterozygous hemoglobin variants, exceeding variant levels of 30% to 40%. Relative bias between the methods near medical decision points of 6.5% and 8.0% (48 and 64 mmol/mol, respectively) achieved the ±5% criterion used by the National Glycohemoglobin Standardization Program.