Progress report on a national quality-control survey of triiodothyronine and thyroxin assay.

In January 1980, a national external quality-control survey was organized to evaluate assays for triiodothyronine (T3) and thyroxin (T4). Currently, about 150 laboratories are involved. Each participant has received and assayed 100 quality-control samples during four periods of about six months each. The average analytical performance achieved by the participants in each six-month period was estimated by computing the average between-laboratory agreement (CVT), the overall average bias, and the average laboratory imprecision. During the 2.5 years of the survey, analytical performance has improved for both assays (CVT decreased from 17.0 to 15.7% for T3 and from 13.1 to 12.7% for T4). Analysis of survey results according to the method/kit used (mean kit bias and kit imprecision for the nine kits most used by participants) showed that the analytical reliability of the T4 assay is generally better than that observed for T3, mainly because of the larger systematic differences among T3 kits.

Radioimmunoassay of unconjugated and total serum estetrol using a 125I-iodinated tracer.

A radioimmunossay (RIA) for the measurement of both unconjugated and total serum estetrol has been developed, using an antiserum to an E4-3-conjugate and a 125I-radioiodinated E4 tracer. Assay of dried ethyl ether extracts was used for the determination of unconjugated E4, while a direct measurement of unextracted hydrolyzed serum in the presence of 0.3% 8-anilino-1-naphthalene sulphonic acid (ANS) proved adequate for total E4. Assay reliability was evaluated and the procedure standardized through a series of tests aimed at assessing accuracy, sensitivity and precision. No steroidal interference was found to practically affect the assay (0.3% estriol cross-reactivity), nor were solvent and sample blanks observed in the case of unconjugated E4. For total E4 assay, the sample blank effects were acceptably overcome by using hydrolyzed male serum and 0.3% ANS, as a standard diluent. An interassay variability amounting to approximately 10 and 6% resulted for unconjugated E4 and total E4 RIA, respectively. A number of serum samples (285 for unconjugated E4, 147 for total E4) randomly collected throughout normal pregnancy were assayed. The unconjugated E4 levels at 15th week and at term were 62.7 +/- 22.6 and 766.5 +/- 208.2 (SD) pg/ml, respectively. Total E4 was about 6--7 times higher than the levels of free E4 and increased 7 times from the 15th week to term.

Direct radioimmunoassay of estradiol in serum extracts.

The feasibility of direct radioimmunoassay of unconjugated estradiol in dried serum extracts in the assessment of ovarian function in the regulation of the menstrual cycle has been investigated. Using an antiserum to an estradiol-6-conjugate, assay reliability was evaluated and the procedure standardized through a series of tests aimed at assessing accuracy, sensitivity, and precision. These included multiple titration with competing steroids, the definition of blank effects from solvent and sample, the assay both of samples to which known amounts of estradiol had been added and of serially diluted samples, and clinical validation using as a reference samples related to well-defined physiological situations. The variability of replicate estimates and the repeatability of the calibration curve were evaluated to obtain information on assay precision and sensitivity. The analytical performances proved to be perfectly adequate for the clinical purposes for which the assay was intended, in terms both of reliability of results and of methodologic practicability.

Measurement by radioimmunoassay of unconjugated estriol in pregnancy serum.

Antisera to estriol 6--onjugates were tested for suitability in the direct radioimmunoassay of unconjugated estriol in extracts of pregnancy serum. Assessment of specificity through titration with competing steroids allowed a selection to be made within the group of antisera. The measurement was then standardized by checking the absence of analytical blank (solvent and sample blank) and the extent of binding variability associated with the use of charcoal-dextran as a separating agent. Validation was made by means of the usual recovery and dilution tests, and by a cross-comparison of analytical data obtained with different antisera; reproducibility of calibration curve and within-and between-assay variability was evaluated under routine conditions. The validity of the clinical information was assessed by assaying 202 samples randomly collected throughout normal pregnancy from the 16th week to term: both trend and levels of unconjugated estriol concentration were found to be in good agreement with the literature data.