Interference in immunoassay is an underestimated problem.

The presence of antibodies in some patients' serum has long been known to be a potential source of interference in immunoassays, as shown by numerous case reports. These often appear after the introduction of a new analyte (e.g. troponin) and then decrease in number as the topic becomes exhausted. This highlights the persistent and intrinsic nature of this problem, despite attempts by the manufacturers to compensate for this source of error. However, an explanation of the immunoanalytical basis underpinning these assays could be more effective in raising awareness than intermittent case reports. In this review we have outlined the use of antibodies as reagents, the factors determining how they bind to antigen(s), and the nature of the immune response in order to explain the insidious and unpredictable nature of this form of interference. Studies on the prevalence of interference have yielded values ranging from 0.05 to more than 2%. However, these figures are analyte- and assay-specific, influenced by the study design, and are not therefore generally applicable. It is also highly likely that figures on prevalence and incidence will worsen in the future because of the wider use of monoclonal antibodies as diagnostic and therapeutic tools. Clinical laboratories should be alert to assay interference from antibodies irrespective of its nature, as immunoassays will remain an indispensable analytical tool, unlikely to be replaced in the foreseeable future by a practical alternative.

Steroids in human intrauterine fluids of early pregnancy.

OBJECTIVE: Little is known of the hormone environment of the developing early human embryo. We have therefore measured selected steroids in the intrauterine fluids of early pregnancy. DESIGN: Measurement of progesterone, 17 alpha-hydroxyprogesterone, oestradiol-17 beta, testosterone, androstenedione, cortisol and dehydroepiandrosterone sulphate in matched samples of coelomic fluid, amniotic fluid and maternal serum collected before pregnancy termination from 12 women between 8 and 12 weeks gestation. RESULTS: Mean concentrations of progesterone, oestradiol-17 beta and 17 alpha-hydroxyprogesterone in coelomic fluid were respectively 20, 6 and 2 times greater than in maternal serum and 8, 13 and 2.6 times those in amniotic fluid. Concentrations of testosterone and androstenedione were highest in maternal serum and lowest in amniotic fluid. Cortisol and dehydroepiandrosterone sulphate were found in intrauterine fluids only at the limit of detection but in normal concentrations in maternal serum. CONCLUSIONS: Coelomic fluid contains relatively high concentrations of progesterone, oestradiol-17 beta and 17 alpha-hydroxyprogesterone which may be synthesized locally. Amniotic fluid contains lower concentrations of steroids (other than progesterone) than are found in coelomic fluid or maternal serum. Free diffusion of steroids across the amnion appears limited. This may constitute a mechanism to protect the embryo from unwanted exposure to biologically active steroids.

A method for the determination of sex hormone binding globulin using Concanavalin A-sepharose.

A binding assay for sex hormone binding globulin (SHBG) has been developed in which SHBG is saturated with tritiated dihydrotestosterone (DHT). Separation of bound and free DHT is achieved by using Concanavalin A-Sepharose as a solid phase matrix. The method is described and its performance, including linearity, imprecision and comparison with other methods, is assessed. The assay is simple and robust and is suitable for analysis of samples of plasma or serum for clinical or research use.

Acne is not associated with abnormal plasma androgens.

Plasma dehydroepiandrosterone sulphate, androstenedione, testosterone (T), dihydrotestosterone (DHT), and sex hormone binding globulin (SHBG) have been measured in 64 females and 26 males aged less than 25 years and with acne vulgaris. Oestradiol was measured in the males. Free T and free DHT were calculated. Acne was graded on three sites and the sebum excretion rate (SER) was measured in most patients. With the possible exception of free DHT, none of the plasma steroids or SHBG correlated with acne severity or with SER. Free DHT in the females showed a possible, but weak, correlation with total acne (r = 0.25, P = 0.07), but comparison with male data showed that this was not causative. The role of androgens in acne is permissive and plasma androgen measurements usually have no place in its management.

Cortisol binding capacity and oestrogen concentrations in maternal and cord plasma in pregnancies with normal and anencephalic fetuses.

The cortisol binding capacity of maternal and cord plasma samples obtained at delivery from fifteen women and their normal infants and from seven women and their anencephalic infants was measured at 4 degrees C by a gel filtration technique. The concentration of oestrogen in these samples was measured by radioimmunoassay. There was no significant difference (t test) between the cortisol binding capacity of peripheral plasma from women with normal infants (1-55 +/- 0-24 mumol/1, mean +/- SD) and from those who delivered anencephalic infants (1-35 +/- 0-30 mumol/1), nor between the cortisol binding capacity of cord plasma from anencephalic infants (0-47 +/- 0-04 mumol/1) and that of normal infants (0-37 +/- 0-10 mumol/1). However, mean oestrogen concentrations in maternal and cord plasma from the pregnancies with an anencephalic fetus were significantly lower (P less than 0-01) than in the corresponding samples from normal pregnancy. It is concluded that oestrogen concentrations in maternal and cord plasma in normal pregnancy at delivery are much greater than those required to account for the increase in plasma cortisol binding capacity. Since plasma cortisol binding capacity in pregnancy with an anencephalic fetus is not diminished, the reduced excretion of corticosteroids relative to normal pregnancy in this condition is unlikely to be due to alterations in cortisol metabolism associated with a lower plasma cortisol binding capacity.

Corticosteroid production by the human foetus: evidence from analysis of urine from women pregnant with a normal or an anencephalic foetus.

The quantities of nine corticosteroids in 24 h urine samples collected by pregnant women (nine with normal foetuses and nine with anencephalic foetuses) were measured after hydrolysis with beta-glucuronidase and separation by paper chromatography. The excretion (mumol/24 h, mean +/- S.D.) of prenanetriol (0-85 +/- 0-17), 3 alpha, 17 alpha-dihydroxy-5 beta-pregnan-20-one (17 alpha-hydroxypregnanolone, 0-55 +/- 0-17), 3 alpha, 17 alpha, 21-trihydroxy-5 beta-pregnan-20-one (tetrahydro-11-deoxycortisol, 0-17 +/- 0-14) and tetrahydrocorticosterone (0-65 +/- 0-26) by women with an anencephalic foetus was significantly lower (P less than 0-01 or less than 0-05) than the excretion of these compounds by women with a normal foetus (pregnanetriol, 2-42 +/- 0-62; 17 alpha-hydroxypregnanolone, 2-72 +/- 0-69; tetrahydro-11-deoxycortisol, 0-56 +/- 0-37; tetrahydrocorticosterone, 1-95 +/- 0-94). These differences suggest that the adrenal of the normal foetus contributes to the quantity of pregnanetriol, 17alpha-hydroxypregnanolone, tetrahydro-11-deoxycortisol and tetrahydrocorticosterone in maternal urine. The excretion of tetrahydrocortisol, tetrahydrocortisone, tetrahydrodeoxycorticosterone, cortol and cortolone were similar in both groups of subjects. No evidence was obtained therefore to indicate the secretion of cortisol or deoxycorticosterone by the foetal zone of the adrenal of the undisturbed human foetus in late gestation.