The authors aimed at developing a simplified method for the measurement of plasma free (unbound) steroids with ultrafiltration using Grace MPS device, and its clinical application. In this method, the movement of free steroids from plasma into ultrafiltrate was monitored with [14C]glucose added to the plasma.
Plasma free steroids (cortisol, testosterone, estradiol and prednisolone) were measured as follows: Plasma was incubated with [14C]glucose (1.2×104dpm/5μl) for 30 min at 37°C. A 0.5ml of aliquot was transferred to MPS device containing a single YMT membrane, and centrifuged at 1100×g, at 37°C, for 30 min in a 45°fixed angle head with the special temperature controller. After centrifugation, [14C] in 30μl of plasma and ultrafiltrate were counted. For the calculation of plasma free level, steroid concentration of ultrafiltrate measured directly by radioimmunoassay (RIA) was multiplied by the ratio of [14C]glucose (dpm) in plasma to [14C]glucose (dpm) in ultrafiltrate. For RIA of cortisol, testosterone and estradiol, commercially available kits were used, and for prednisolone, anti-prednisolone antibody developed in our labolatory was utilized.
Since plasma free cortisol level showed a parallel increase with rise in temperature, strict temperature control during ultracentrifugation was required. On the other hand, the duration of centrifugation and the sample volume applied to MPS did not show any significant effect on the estimated values. Intraassay and interassay variations were 4.4% and 6.1% in free cortisol, 6.3% and 8.7% in free testosterone, 8.5% and 9.2% in free estradiol, and 8.8% and 9.9% in free prednisolone, respectively.
The correlations between the plasma free steroid levels obtained by ultrafiltration (y) and equilibrium dialysis (x) were as follows, respectively: free cortisol; y=1.16x+0.017(r=0.95, n=10), free testosterone; y=1.17x-0.027(r=0.92, n=10), free estradiol; y=1.33x+8.55(r=0.98, n=12), free prednisolone; y=1.03x+1.00(r= 0.98, n=7).
The mean plasma free cortisol levels and percent free fractions (%FF) were 1.16±0.40(±SD)μg/dl and 10.9±3.0% in 10 control patients, 4.4±1.6μg/dl and 15.6±3.3% in 8 patients with Cushing's syndrome, and 1.45±0.48μg/dl and 5.9±2.4% in 11 normal pregnant women (10-40 weeks of pregnancy), respectively. The mean plasma levels of free testosterone and %FF were 0.16±0.06 ng/ml and 3.6±1.1% in 10 male controls (27±3 yrs.), and 0.033±0.004 ng/ml and 6.4±1.9% in 5 patients with polycystic ovary syndrome, 0.21 ng/ml and 5.1% in a patient with mixed gonadal dysgenesis, and 0.14 ng/ml and 2.8% in a patient with testicular feminization syndrome, respectively. In each patient with Klinefelter's syndrome, panhypopituitarism (male) and Kallmann's syndrome (male), the values were 0.033, 0.021 and 0.008 ng/ml, and 2.6, 3.5 and 5.8%, respectively. In 12 normal pregnant women (10-40 weeks of pregnancy), the plasma free estradiol level and %FF were in the range of 3.1-159 pg/ml and 0.63-5.1%, respectively. Both values tended to increase in proportion to the age of gestation in weeks (r=0.93 and r=0.95, respectively). These values fell within the range of published values for each steroid by various methods. The plasma level of free prednisolone in the healthy volunteers were in parallel with those of total prednisolone observed in our previous report.
In conclusion, the presented method has the advantage of rapid and simple separation of free and bound steroids as well as of a large sample number in a single assay, and it is practical for clinical investigation of free steroids.
