Search

Immunoassays are used daily by clinical endocrinologists to refute or confirm a diagnosis, or to follow up the course of a treatment. Immunoassays have become increasingly sensitive, specific and reproducible, so that clinicians have great confidence in their results. However, they do not always yield correct results because interferences still occur. In the first part of this review, immunoassay interferences are described and their mechanisms explained: matrix effects, specificity defaults, interferences from antibodies and binding proteins, or the hook effect. The hormones most frequently concerned by these pitfalls are reviewed, even as strategies used to prevent and minimize interferences. In the second part, the consequences of failure to recognize an interference are reviewed: unjustified clinical decisions (wrong diagnosis, treatment, surgery) and, more exceptionally, publication in the scientific literature of erroneous results. Clinicians may sometimes inform laboratories of an increased risk of interference. In general, clinicians are in the best position to fully validate test results; because only they know all the medical data about their patients. If there is any suspicion of discrepancy between clinical and laboratory data, they must alert the laboratory. This is the only way to better detect and, if possible, eliminate interferences and their regrettable outcomes.

There are numerous etiologies of hyperprolactinemia, a common reason for consultation. Diagnostic measures must be capable of identifying the tumors, the most frequent of which are prolactin adenomas. Hypothalamic-pituitary MRI is the reference morphological examination. In clinical practice, it is usually performed very early, following the discovery of increased plasma concentrations of PRL. This approach is warranted for marked increase in PRL in the absence of drugs with hyperprolactinemic effects (>10 x upper limit of normal) since a diagnosis of PRL adenoma is extremely likely under such circumstances. When hyperprolactinemia is moderate, which is the most common finding in practice, all etiologies are possible in theory and it is important to follow a rational diagnostic plan (history-taking to identify use of any drugs with hyperprolactinemic effects paying attention to renal and hepatic history, investigation for endocrine diseases occasionally associated with hyperprolactinemia such as hypothyroidism or polycystic ovary syndrome (PCOS), confirmation of hyperprolactinemia by a second assay when the initial level is less than five times the upper normal limit, pregnancy testing for women of childbearing age) in order to rule out all non-tumoral causes of hyperprolactinemia before proceeding with imaging. Absence of any consequences of hyperprolactinemia on gonadic function or the existence of a concomitant disease that could account for the clinical signs, demonstration of wide variations in PRL from one assay to another in a single patient could prompt screening for macroprolactinemia before MRI is ordered. Macroprolactinoma could also occur in the case of normal or doubtful MRI or discrepancy in response to medical or surgical treatment. T1- and T2-weighted coronal sections (with or without T1 after gadolinium injection) are generally sufficient for diagnosis of microprolactinoma. Dynamic tests may be useful if MRI is normal or unclear. Gadolinium injection with sagittal and axial sections is essential for examination of large lesions. In this case, when the increase of PRL is moderate (<150 mg/ml), a non-lactotropic lesion may be suspected without misdiagnosing a hook effect. Careful analysis of the images allows differentiation between tumoral lesions and pituitary hyperplasia.

Thyroid hormone (TH) homeostasis depends on peripheral activation and inactivation of iodothyronines by selenoenzymes of the deiodinase (Dio) family. We genetically inactivated hepatic selenoenzyme expression, including Dio1, in order to determine the contribution of hepatic Dio to circulating TH levels. Serum levels of TSH, total T(4), and total T(3) were not different from controls. We measured Dio1 and Dio2 in kidney, skeletal muscle, heart, brown adipose tissue, and brain, but did not find compensatory up-regulation in these tissues. Finally, we determined expression in the liver of the following T(3) target genes: Spot14, alpha-glycerophosphate dehydrogenase (alphaGPD), and malic enzyme (ME). On the transcript level, both Spot14 and alphaGPD were reduced in Dio-deficient liver to about 60-70% of controls. However, mRNA and activity of ME were significantly increased in the same mice. Together, our results indicate that hepatic Dio1 activity is not absolutely required to sustain the euthyroid state in mice.

Background: The Elecsys prolactin (PRL) assay reacts more strongly with macroprolactin than the Centaur PRL assay. We evaluated Elecsys direct and free PRL measurements vs. Centaur direct PRL measurements, in sera with and without macroprolactin., Methods: PRL was measured using Elecsys and Centaur direct assays and the Elecsys assay in the supernatant obtained after PEG precipitation (free PRL) in 34 sera from 34 hyperprolactinemic male subjects (Elecsys direct PRL>434 mIU/l) classified, according to the PRL recovery after PEG precipitation, as: negative, i.e. without predominant macroprolactin (recovery %>50, n=12), positive (recovery %<40, n=18) or indeterminate (n=4)., Results: The positive bias between Elecsys direct and Centaur PRL results was clearly influenced by the presence of macroprolactin and the mean bias between Elecsys free and Centaur prolactin values was near zero in the negative and positive groups. Among 14 patients from the positive group presenting clinical conditions possibly ascribable to hyperprolactinemia, Elecsys free and Centaur PRL levels were normal in seven and increased in five. In the negative and positive groups considered together, Elecsys free PRL agreed well with Centaur prolactin (28/30). The poor concordance observed for the indeterminate samples underlined the heterogeneity of macroprolactin., Conclusion: Elecsys free PRL determination can be used to reduce the marked influence of macroprolactin in this assay.

Introduction: Patients with nonthyroidal disease frequently exhibit abnormal thyroid function tests; this is referred to as euthyroid sick syndrome. The clinical significance of this syndrome is unknown: abnormal endocrine reaction with reduced triiodothyronine (T3) at the tissue level, or adaptation to stress protecting the body against exaggerated catabolism., Current Knowledge and Key Points: Recent advances in the underlying mechanisms concern the role of deiodinase and of the transport of thyroid hormone in tissues. Various factors acting on deiodinase or on transport system, such as medications and nutritional factors, have been implicated. Considerable interest has raised concerning the role of cytokines. Some cytokines may act at every level of the thyrotropic axis, but their real action in vivo remains unclear. Nutritional factors have a great impact on thyroid hormone metabolism, but the mechanism of the decrease in T3 induced by starvation is not identified. The role of the decrease in type I hepatic deiodinase has been recently challenged., Future Prospects and Projects: Despite its complexity, euthyroid sick syndrome is a model for the study of thyroid hormone metabolism regulation. Characterisation of the thyroid hormone transport proteins will lead to significant advances in the understanding of the syndrome.

The consequences of aging function are appreciated in various ways. In fact most of the thyroid hormone abnormalities described in the elderly are secondary to associated disease or treatments. Health and nutritional status seem to be more determinant than age. However, there is an increased prevalence of dysthyroidism in the elderly. Symptoms may be misleading and induce severe general repercussions. Mild or occult forms of dysthyroidism characterized by abnormal TSH levels with normal free thyroid hormone levels are not uncommon. Their clinical significance and their prognosis without treatment are still unresolved. The frequency of dystrhyroidism during old age and the possibility of a specific treatment may be in favour of a screening by a systematic TSH determination in elderly patients.

Amiodarone (A) treatment alters the levels of thyroid hormones. We investigated whether new hormonal assays are also altered by this drug. Thyroid function was determined in 21 patients chronically treated with A and in 30 controls. TSH was determined with a third generation assay. Free T3 and Free T4 levels were determined by 5 different immunoassays. Equilibrium dialysis (E.D.) was considered as the reference assay for FT3 and FT4. With this method FT3 is diminished, reverse T3 ans FT4 are increased whereas TSH remained normal. Only FT3 determined by an assay using a labelled monoclonal antibody (MAB) appears not modified by A. Other methods (chromatography, chemiluminescence and radioisotopic) give results in agreement with E.D. Some of these alterations may be explained by a marked increase in plasma levels of non esterified fatty acids. Hormonal changes induced by A are typically with all but one assays. Whatever the method used the determination of TSH remains necessary to avoid misinterpretation of the thyroid function tests in A-treated patients.

The concentration-specific smallest detectable difference (D) between two concentrations (i.e., the sensitivity) was determined from the results of within-run and run-to-run reproducibility studies of a third-generation thyrotropin assay. Only low concentrations were considered (range 0-0.3 m U/l). Assays were performed in singlet or in duplicate. Findings confirmed the value of the assay: the analytic detection limit approximated 0.005 mU/l and the functional detection limit was around 0.01 mU/l. Expressed as a percentage of the concentration, D% was under 20% for levels of less than 0.02 mU/l in within-run studies or of less than 0.03 mU/l in run-to-run studies. With concentrations in excess of 0.05 mU/l, D% was under 10% in within-run studies and slightly above 10% in run-to-run studies. These new detection thresholds for this assay should be taken into account when interpreting clinical results from a single run (TRH test) or from several runs (hormone replacement therapy or suppression test).

The performances of a new "3rd generation" chemoluminescence TSH assay (TSH ICMA) with a functional sensitivity of 0.005 mU/l were compared with those of an "ultrasensitive" TSH immunoradiometric assay (TSH IRMA) in a series of patients characterised by a TSH IRMA less than 0.20 mU/l and normal free thyroxin (T4 L) and triiodothyronine (T3 L) concentrations. The 95% cut-off value for hyperthyroidism was 0.03 for TSH ICMA and 0.05 for TSH IRMA. In a first group of 41 subjects undergoing Tc99m thyroid scan, images of multifocal increased uptake or toxic adenoma were associated with a lower TSH ICMA than in patients with a normal isotope scan. TSH ICMA was also lower than TSH IRMA (p < 0.01). At the cut-off value of 0.03 mU/l, the specificity of TSH ICMA was higher than that of TSH IRMA, but the sensitivity were identical. In a second group of 36 patients with severe non-thyroid diseases, TSH ICMA was lower than the cut-off value for hyperthyroidism in 30% of cases, while TSH IRMA was lower than the cut-off value in 40% of cases. A satisfactory concordance was observed between the two methods. In conclusion, the two TSH assays, IRMA and ICMA, provide globally comparable information in subjects with a low TSH and normal T4 L and T3 L. However, the better specificity of TSH ICMA and a smaller overlap with the frank hyperthyroid zone in patients with non-thyroid disease argue in favour of the use of this new assay method.

The comparison of the results of two insulin assays, a RIA using a polyclonal antibody and a EMIA using two monoclonal antibodies in blood samples collected during an oral glucose load in normal, obese, glucose intolerant and type II diabetic patients indicated that hyperinsulinemia of obese and glucose intolerant patients is not associated to increased proinsulin/insulin ratio.

In non thyroid illnesses the degree of serum FT4 levels perturbation may be variable according to the assay used. The performance of a new FT4 immunoenzymatic assay (Enzymmum test, BMF) was evaluated in 78 euthyroid controls and in 99 sick patients with renal insufficiency, severe diabetes, hypoalbuminemia, severe general disease or under heparinate treatment. Results were compared to two RIA assays: by immunoextraction (CA2), known to be little disturbed by hypoalbuminemia and considered as a reference method, and by a labelled antibody (MAB, Amerlex) easier to use. In controls mean values and T4 L confidence intervalls were comparable with the three assays. In patients FT4 levels were diminished in case of renal insufficiency but the mean values obtained by BMF and CA2 did not differ. FT4 levels of heparinate treated patients were elevated with BMF and CA2. Generally patients are better classified as euthyroid with the CA2 assay (5.6% misclassification) than with BMF (11.2%) or MAB (12.4%), particularly in case of serious disease. A good linear correlation was found between the three methods. With the new BMF assay results were grossly comparable to those obtained by other assays. However the determination of FT4 levels alone appeared no sufficient to characterize the patients thyroid status.

Treatment of primary hypothyroidism requires a long-term hormonal therapy with levothyroxine which restores a normal TSH value. In a series of 192 patients followed up for at least 3 years we tried to evaluate the variability of maintenance dosage and to analyse the factors predictive of replacement dosage. The L-T4 dosage varied from one individual to the other, from 0.6 to 3.2 micrograms/kg. The mean maintenance dosage (118 +/- 38 micrograms/day) was similar in men and women weight for weight, but it was higher in subjects aged under 60 and in those with Hashimoto's or congenital hypothyroidism. There is a positive linear relation between the equilibrium dose and the TSH value (n = 0.84; p < 0.001) and a negative relation with the L-T4 concentration (n = 0.71; p < 0.001). The yearly variation of dosage in each individual was small (about 3.5%) and of the same order as the variation of weight. We conclude that the L-T4 maintenance dosage depends on the cause of hypothyroidism, on the importance of biochemical abnormalities and on the patient's age, sex and weight. In practice, however, the necessary dosage is difficult to predict. In the equilibrium phase the need for replacements is remarkably stable. A yearly assay of baseline TSH seems to be sufficient to evaluate the quality of hormonal compensation in the absence of intercurrent events.

The myocardial infarction (M.I.) constitutes an exemplary acute severe affection able to modifie hormonal concentrations. The total and unbound thyroid hormones, reverse T3 (rt3), TSH, and cortisolemia were determined in 24 patients during a period of 21 days in order to compare them to different markers of severity of MI. The initial phase of the disease is characterized by low concentrations of total and free T3 and high concentrations of rT3 combined with more often than not normal total and free T4 and TSH values contrasting with an increase in cortisol levels. The abnormalities were more pronounced the day after admission and then progressively amend. There are several statistic relationship between the marker of severity of MI and thyroid hormones. In the same way total T3 is all the more decreased especially since myoglobin, CPK-MB, ST amplitude and ventricle ejection fraction are more disturbed. Severe forms of MI induces a pseudo central thyroid insufficiency with low T3, low T4 and a tendency to TSH decrease. Total T3 blood levels may usefully contribute to the elaboration of an MI severity index.

Insulin secretion is assessed by simultaneous radioimmunological assay of insulin (R.I.I.) and C-peptide (R.I.C.P.) levels under basal conditions, and after stimulation by oral or intravenous glucose administration. Subjects with abnormal glucose tolerance demonstrate an increase in the ratios R.I.C.P. divided by glucose and R.I.C.P. divided by R.I.I. after fasting, increase in R.I.I. and R.I.C.P. occurring later and lasting longer after glucose loading. These anomalies are observed in both obese subjects and those with normal body weights. Simultaneous determination of R.I.I. and R.I.C.P. levels appears of value in a limited number of cases where glucose loading tests do not supply precise information on the quality of glucose tolerance. Obese subjects, in whom abnormal hyperglycemia levels provoked by oral glucose are observed, but in whom the peripheral glucose assimilation coefficient is normal, can be considered to be non-diabetic as shown by the levels of R.I.I. and R.I.C.P. and more particularly by the molar ratio R.I.C.P. divided by R.I.I.

Serum samples from 21,342 patients undergoing evaluation of thyroid status were screened for familial dysalbuminemic hyperthyroxinemia (FDH) using a specific test based on the measure of charcoal uptake of 125I thyroxine (T4) from serum diluted 1:100 with addition of unlabelled 10(-6) M T4. We found 17 cases of FDH: a higher incidence (8:10,000) than previously reported in the general population (1:10,000). The results of thyroid function tests of patients with FDH are presented: total T4 concentration is increased in only 14 subjects; thyrotropin and free T4 measured by an immunoextraction method are the most useful assays to evaluate the clinical status of these patients.

Changes in total and free thyroid hormones, reverse triiodothyronine (rT3) and TSH were followed in 64 patients undergoing surgery and divided into three groups according to nature of surgery. Group I patients underwent extraperitoneal surgery, group II mild and group III heavy intraperitoneal surgery. A significant fall in total and free triiodothyronine (T3), a rise in rT3, free thyroxine (FT4) and in TSH were noted in each group after surgery, while total T4 levels remained unchanged. There was no correlation between the change in thyroid hormones and TSH. These changes are more marked in groups II and III than in group I. The rT3/T3 ratio was more elevated after surgery in group III than in group II. Thus the severity of the stress induced by surgery may interact with the mechanism of peripheral T4 conversion. The rise in FT4 may be interpreted as a consequence of the presence of an inhibitor of thyroid-hormonal binding to serum proteins. The slight increase in TSH levels seems rather due to stress than to a feed-back adaptation. The ratio rT3/T3 appears as an index of severity and prognosis.

Salivary cortisol levels reflect the biologically active "free" fraction of blood cortisol. The authors describe the results obtained with the aim of a radio-immunoassay commercial serum cortisol kit, without prealable extraction in different physiological and pathological situations. Salivary cortisol determination appears performant both in nycthemeral studies and in stimulation or freination tests.

A new ultrasensitive TSH immunoradiometric assay (IRMA) using two monoclonal antibodies is now able to distinguish between euthyroid and hyperthyroid patients. The aim of this study was to compare data given by ultrasensitive basal TSH (IRMA) and by the response of TSH to TRH test considered until now as the more reliable test in case of mild or atypical hyperthyroidism. Basal plasma TSH levels were determined in euthyroid (n = 80), hyperthyroid (n = 30), hypothyroid (n = 14) and pituitary deficient patients (n = 8) before and 30 minutes after a TRH test (250 micrograms i.v.). A close linear correlation was found between basal and post-stimulative TSH levels. Normal TSH response ranged from 2 to 22 uU/ml. The sensibility and the specificity of these two parameters appeared comparable in the case of primary dysthyroidism; on the contrary basal TSH levels were not sufficient for the diagnosis of central hypothyroidism. In conclusion, excepted for pituitary deficiency, basal plasma TSH (IRMA) levels are accurate and sufficient in the evaluation of the thyroid function and make the TRH-test useless.

Total serum thyroxine (T4), total serum triidothyronine (T3), thyrotropin (TSH) and TRH induced TSH release have been measured in 50 clinically euthyroid men with alcoolic cirrhosis and compared to 20 age matched control men in order to determine the thyroid status in cirrhosis. Free serum thyroxin (FT4), free triiodothyronine (FT3) and reverse T3 (rT3), were measured in 34 patients. In these patients, a clinical and biological index was devised to score the severity of the disease. Finally, thyroxine-binding-globulin (TBG) levels were measured in 20 patients. In cirrhotics, the mean total T4 level is normal but the mean T3 level is markedly reduced: FT4 is slightly raised and FT3 is decreased. Serum rT3 levels are often very high and correlated in T3. Basal TSH concentration were slightly significantly higher than normal; the mean magnitude of TSH responses to TRH is comparable to controls but the individual responses are very variable, insufficient or exaggerated, and non correlated with other thyroid function tests. The mean TBG level is also normal and non correlated with total free thyroid hormones. Correlation of thyroid function tests with liver function showed significant correlations between T3, FT3, rT3 or TSH and serum albumin concentration and particularly between T3, rT3 and the clinical and biological index. So the ratio rT3/T3 may be proposed as a valuable index of prognosis and severity of the disease. In conclusion, in alcoholic cirrhosis, thyroid function and regulation are characterized by normal T4 and low T3 levels related to reduced extrathyroidal T4 and T3 conversion levels - and thus maintain the euthyroid state - and by a hypothalamo-pituitary dysfunction. These alterations are related to the degree of liver dysfunction.