Comparison of radial immunodiffusion and alkaline cellulose acetate electrophoresis for quantitating elevated levels of fetal hemoglobin (HbF): application to evaluating patients with sickle cell disease treated with hydroxyurea.

Radial immunodiffusion (RID), alkaline cellulose acetate electrophoresis, and high-performance liquid chromatography (HPLC) were compared for quantitating the elevated (> 10%) level of fetal hemoglobin (HbF) found in the red blood cells of sickle cell disease patients undergoing treatment with hydroxyurea. HPLC- and electrophoresis-determined values were comparable. The RID-determined values were higher, in many cases twofold higher. False high HbF values would be misleading in assessing the effectiveness of hydroxyurea therapy in sickle cell disease patients. We subsequently initiated an examination of the variation in HbF values due to the use of different HbF radial immunodiffusion QUIPlates and different positions within a single plate in an attempt to determine the cause of these discrepancies. Within-run precision studies indicated that significantly different size precipitin rings were obtained depending upon which area of the plate the hemolysate containing antigen (HbF) was applied. A common feature associated with poor precision plates was a marked difference in degree of coloration of gel throughout the plate. Spuriously high HF concentrations were obtained with antigen (HbF) placed in wells located in the lighter colored gel area while antigen placed in wells in the darker colored area of the agarose gel bed were more in agreement with the electrophoretically determined HbF concentrations. The variation in HbF values was significantly greater in the diluted (HbF QUIPlate Diluent) samples than in the neat samples even on plates of uniform gel coloration. As a result of this study, we will continue to monitor high HbF levels by densitometry following alkaline cellulose acetate electrophoresis.