Liquid chromatography and radioimmunoassay compared for determination of cortisol and corticosterone in plasma after a dexamethasone suppression test.

We developed sensitive, specific "high-performance" liquid chromatography (HPLC) for determining suppressed cortisol and corticosterone in human plasma and compared its efficacy with that of conventional radioimmunoassay (RIA) at concentrations in the nanogram per liter range. Steroids from a 0.5-mL aliquot of plasma were extracted by rapid-flow fractionation, with diethyl ether as mobile-phase solvent, diatomaceous earth granules as stationary-support material. Analytical recovery of the steroids approached 100%. Concentrations in plasma were determined from peak-height ratio calibration (dexamethasone internal standard). The analytical column contained silica gel and the solvent system was water/methanol/dichloromethane/n-hexane (0.1/3/30/66.9 by vol). We could measure the steroids before and 20 h after oral administration of 0.5 mg of dexamethasone. The detection limit was 300 ng per liter of plasma for corticosterone, 500 ng/L for cortisol, with CVs of less than 4%. Determining corticosterone after administration of dexamethasone, in four of 20 such samples we could determine concentrations greater than 300 ng/L; the others contained corticosterone between 100 and 300 ng/L, but these values could not be certified analytically. Mean concentrations of these hormones as determined by RIA substantially exceeded those by HPLC. Some cross reactions in RIA could not be considered negligible in spite of pre-column treatment of the extracts.