The use of human glutathione S-transferase A1 in the detection of cystic fibrosis liver disease

OBJECTIVE

To determine the value of serum human glutathione S-transferase A1 (hGST A1) in the detection of cystic fibrosis liver disease (CFLD).

METHODS

Sixty-three children (aged 0.5-16 years) with cystic fibrosis (CF) were screened prospectively for evidence of hepatobiliary abnormalities between February 1993 and February 1996. Comparison was made between clinical examination, abdominal ultrasonic scan, measurement of conventional liver enzymes (LFTs) and serum hGST A1 concentration in the detection of hepatobiliary abnormalities in children with CF.

RESULTS

The 5-95% concentration of serum hGST A1 was 1.7-4.27 micrograms L-1 for the control group. The hGST A1 levels in the CF patients were significantly higher than in the non-CF group. Thirty-eight (60%) children had detectable hepatobiliary abnormalities. Ultrasound scanning detected the highest number of abnormalities (41%), followed by hGST A1 (30%). The presence of clinical liver disease was found in 19% of the children. The estimated sensitivities of detecting CFLD by clinical method, ultrasound scan, serum hGST A1, and LFTs would be 32%, 68%, 50% and 16%, respectively.

CONCLUSIONS

Serum hGST A1 measurement increases the sensitivity of detecting hepatic abnormalities when included with clinical and ultrasound evaluation although, in some cases with advanced liver disease, serum hGST A1 may be normal. Conventional liver enzyme tests add little information in the detection of CF liver disease.

Effect of propylthiouracil-induced hypothyroidism on the onset of skeletal muscle necrosis in dystrophin-deficient mdx mice

1. Duchenne and Becker muscular dystrophies are X-linked disorders caused by defects in muscle dystrophin. The mdx mouse is an animal model for Duchenne muscular dystrophy which has a point mutation in the dystrophin gene, resulting in little (<3%) or no expression of dystrophin in muscle. Mdx mice show a characteristic pattern of muscle necrosis and regeneration. Muscles are normal until the third postnatal week when widespread necrosis commences. This is followed by muscle regeneration, with the persistence of centrally nucleated fibres. 2. This work has examined the hypothesis that the onset of this muscle necrosis is associated with postnatal maturation of the thyroid endocrine system and that pharmacological inhibition of thyroid hormone synthesis delays the onset of muscle necrosis. 3. Serum T4 and T3 concentrations of mice were found to rise immediately before the onset of muscle necrosis in the mdx mouse, and induction of hypothyroidism by treatment of animals with propylthiouracil was found to delay the onset of muscle necrosis. 4. The results provide the first demonstration of experimental delay of muscle necrosis by manipulation of the endocrine system in muscle lacking dystrophin, and provide a novel insight into the way in which a lack of dystrophin interacts with postnatal development to precipitate muscle necrosis in the mdx mouse.

Identification of a 57-kilodalton selenoprotein in human thyrocytes as thioredoxin reductase and evidence that its expression is regulated through the calcium-phosphoinositol signaling pathway

Human thyrocytes incubated with the phorbol ester, phorbol 12-myristate 13-acetate (PMA; 10(-5)-10(-8) mol/L) and the calcium ionophore A23187 (10(-5)-10(-8) mol/L) showed a marked increase in the expression of a 57-kDa selenoprotein identified as thioredoxin reductase (TR). After the addition of A23187 with PMA, a significant induction in TR expression was observed after 6 h, with maximal induction occurring by 24 h. The addition of 8-bromo-cAMP (10(-4) mol/L) or TSH (10 U/L) alone had no effect on TR expression, nor did these agents influence the induction of TR brought about by the addition of A23187 and PMA. These data show that the calcium-phosphoinositol second messenger cascade that controls hydrogen peroxide generation in the human thyrocyte is also an important stimulator of TR expression. The role of TR in the thyrocyte is unclear, but the selenoenzyme has a high capacity to detoxify compounds, such as hydrogen peroxide and lipid hydroperoxides, that are produced in high concentration during thyroid hormone synthesis.

Differential expression of selenoproteins by human skin cells and protection by selenium from UVB-radiation-induced cell death

The generation of reactive oxygen species has been implicated as part of the mechanism responsible for UVB-radiation-induced skin damage. In mice, evidence suggests that increased dietary selenium intake may protect skin from many of the harmful effects of UVB radiation. We sought to determine the selenoprotein profile of cultured human skin cells and whether selenium supplementation could protect keratinocytes and melanocytes from the lethal effects of UVB radiation. Labelling experiments using [75Se]selenite showed qualitative and quantitative differences in selenoprotein expression by human fibroblasts, keratinocytes and melanocytes. This was most noticeable for thioredoxin reductase (60 kDa) and phospholipid glutathione peroxidase (21 kDa); these proteins were identified by Western blotting. Despite these differences, we found that a 24 h preincubation with sodium selenite or selenomethionine protected both cultured human keratinocytes and melanocytes from UVB-induced cell death. With primary keratinocytes, the greatest reduction in cell death was found with 10 nM sodium selenite (79% cell death reduced to 21.7%; P<0.01) and with 50 nM selenomethionine (79% cell death reduced to 13.2%; P<0.01). Protection could be obtained with concentrations as low as 1 nM with sodium selenite and 10 nM with selenomethionine. When selenium was added after UVB radiation, little protection could be achieved, with cell death only being reduced from 88.5% to about 50% with both compounds. In all of the experiments sodium selenite was more potent than selenomethionine at providing protection from UVB radiation.

Iodine deficiency in cattle: compensatory changes in thyroidal selenoenzymes

The trace elements selenium and iodine are both essential for normal thyroid hormone metabolism. To investigate the relationships between these functions, heifers were maintained on iodine-deficient or iodine-sufficient diets from mid pregnancy to term. In these heifers and their offspring the interrelationship between iodine and selenium was apparent with the preferential 10- to 12-fold induction of the selenoenzyme, thyroidal type I, selenium-containing iodothyronine deiodinase activity by iodine deficiency. This was accompanied by two- to four-fold increases in cytosolic glutathione peroxidase activity, probably reflecting increased oxidative activity and metabolism in the thyroid gland in response to iodine deficiency. The above selenoenzyme activities were not affected in liver, kidney, pituitary and brain by iodine deficiency. The results are consistent with a critical role for selenium in both the normal function of cattle thyroid and key enzymes to compensate for the effects of iodine deficiency.

Selenoprotein expression and brain development in preweanling selenium- and iodine-deficient rats

Selenium deficiency causes further impairment of thyroid hormone metabolism in iodine-deficient rats and therefore could have a role in the aetiology of both myxoedematous and neurological cretinism in humans. Thyroidal type I iodothyronine deiodinase (ID-I), cytosolic glutathione peroxidase and phospholipid hydroperoxide glutathione peroxidase activities were increased in iodine-deficient adult rats and their offspring at 11 days of age. Thyroidal ID-I activity was unchanged and thyroidal cytosolic glutathione peroxidase activity was decreased by more than 75% by combined selenium and iodine deficiency in 11-day-old rats, indicating that, while the thyroid retained an ability to produce 3,3',5-triiodothyronine (T3), the gland was probably more susceptible to peroxidative damage caused by increased hydrogen peroxide concentrations driven by increased thyrotrophin. Thyroidal atrophy, common in myxoedematous cretinism, did not occur in iodine- or selenium and iodine-deficient rat pups. Iodine deficiency increased brain type II iodothyronine deiodinase activity 1.5-fold in 4-day-old rats and 3-fold in 11-day-old rats, regardless of selenium status. Thus rats were able to activate compensatory mechanisms in brain that would maintain T3 concentrations in selenium and iodine deficiencies. Surprisingly, however, selenium deficiency had a greater effect than iodine deficiency on markers of brain development in rat pups. Expression of the brain-derived neurotrophic factor (BDNF) mRNA was decreased in selenium deficiency in 4- and 11-day-old pups and in combined selenium and iodine deficiency in 4-day-old pups. Iodine deficiency caused an increase in BDNF expression in 11-day-old pups but had no effect on 4-day-old pups. Myelin basic protein mRNA expression in brain was decreased by combined selenium and iodine deficiency in 11-day-old rats.

Mutational analysis of the thyrotropin receptor gene in sporadic and familial feline thyrotoxicosis

The characterization of a spontaneous animal model equivalent to a human form of thyrotoxicosis would provide a useful resource for the investigation of the human disorder. Feline thyrotoxicosis is the only common form of hyperthyroidism found in domestic or laboratory animals, but its etiopathogenesis remains poorly defined. We have used the polymerase chain reaction (PCR) to amplify codons 480-640 of the previously uncharacterized feline thyrotropin receptor (TSHR) gene, and have determined the DNA sequence in this transmembrane domain region. We have analyzed single stranded conformational polymorphisms in thyroid DNA from 11 sporadic cases of feline thyrotoxicosis and leukocyte DNA from two cases of familial feline thyrotoxicosis. We have also determined the DNA sequence of this region of the TSHR in five of the cases of sporadic feline thyrotoxicosis and the two familial thyrotoxic cats. The normal feline TSHR sequence between codons 480-640 is highly homologous to that of other mammalian TSHRs, with 95%, 92%, and 90% amino acid identity between the feline receptor and canine, human, and bovine TSHRs, respectively. Thyroid gland DNA from 11 cats with sporadic thyrotoxicosis did not have mutations in this region of the TSHR gene. Leukocyte DNA from two littermates with familial feline thyrotoxicosis did not harbor mutations of this region of the TSHR gene. These studies suggest that TSHR gene mutations are not a common cause of feline thyrotoxicosis.

Selenium and iodine deficiencies: effects on brain and brown adipose tissue selenoenzyme activity and expression

Adequate dietary iodine supplies and thyroid hormones are needed for the development of the central nervous system (CNS) and brown adipose tissue (BAT) function. Decreases in plasma thyroxine (T4) concentrations may increase the requirement for the selenoenzymes types I and II iodothyronine deiodinase (ID-I and ID-II) in the brain and ID-II in BAT to protect against any fall in intracellular 3,3',5 tri-iodothyronine (T3) concentrations in these organs. We have therefore investigated selenoenzyme activity and expression and some developmental markers in brain and BAT of second generation selenium- and iodine-deficient rats. Despite substantial alterations in plasma thyroid hormone concentrations and thyroidal and hepatic selenoprotein expression in selenium and iodine deficiencies, ID-I, cytosolic glutathione peroxidase (cGSHPx) and phospholipid hydroperoxide glutathione peroxidase (phGSHPx) activities and expression remained relatively constant in most brain regions studied. Additionally, brain and pituitary ID-II activities were increased in iodine deficiency regardless of selenium status. This can help maintain tissue T3 concentrations in hypothyroidism. Consistent with this, no significant effects of iodine or selenium deficiency on the development of the brain were observed, as assessed by the activities of marker enzymes. In contrast, BAT from selenium- and iodine deficient rats had impaired thyroid hormone metabolism and less uncoupling protein than in tissue from selenium- and iodine-supplemented animals. Thus, the effects of selenium and iodine deficiency on the brain are limited due to the activation of the compensatory mechanisms but these mechanisms are less effective in BAT.

Selenium and iodine deficiencies and selenoprotein function

This paper reviews some recent findings on the interactions between selenium deficiency and iodine deficiency. Both micronutrients can control the levels of selenoprotein mRNAs, particularly in the thyroid and brain. When selenium and iodine supplies are limiting the compensatory mechanisms work to minimise adverse effects on thyroid hormone metabolism and thus neurological development. The mechanisms for regulation of selenoproteins in selenium and iodine deficiency are however very tissue-specific. For example, unlike the brain and thyroid, brown adipose tissue is unable to retain selenoproteins in selenium and iodine deficiency and is therefore at greater risk from injurious effects of the deficiencies.

Effect of sevoflurane anaesthesia on plasma concentrations of glutathione S-transferase

To assess the effect of sevoflurane anaesthesia on hepatocellular integrity, we measured plasma concentrations of glutathione S-transferase (GST) before anaesthesia and 1, 3, 6 and 24 h after the end of anaesthesia in 41 healthy, Japanese patients undergoing elective, body surface surgery. Sevoflurane (approximately 1.0 MAC) was delivered in 50-66% nitrous oxide in oxygen via a circle system, with a fresh gas flow of 6 litre min-1. Ventilation was spontaneous in all patients. Mean duration of anaesthesia was 101 min. Concentrations of GST increased significantly 1 h after the end of anaesthesia (P = 0.0075), but this was not significantly different from preoperative concentrations at 3, 6 and 24 h. Three patients developed a large secondary increase in GST concentrations at 24 h. The increase observed at 1 h was probably a result of reduced total liver blood flow; the mechanism for the secondary increase at 24 h is unclear but the possibility that products of sevoflurane biotransformation are responsible cannot be excluded.

Selenoenzyme expression in thyroid and liver of second generation selenium- and iodine-deficient rats

The stimulation of thyroid hormone synthesis in iodine deficiency may increase the requirement for the selenoproteins which are involved in thyroid hormone synthesis in the thyroid gland. Selenoenzyme activity and expression were investigated in the thyroid and liver of second generation selenium-and/or iodine-deficient rats. Selenium deficiency caused substantial decreases in hepatic selenium-containing type I iodothyronine deiodinase (ID-I) and cytosolic glutathione peroxidase (cGSHPx) activities and mRNA abundances, but phospholipid hydroperoxide glutathione peroxidase (phGSHPx) activity was only 55% of selenium-supplemented control levels, despite the absence of change in its mRNA abundance. Selenoenzyme mRNA concentrations were maintained at control levels in thyroid glands from the selenium-deficient rat pups. Despite this, a differential effect was observed in selenoenzyme activities: ID-I activity was decreased to 61%, cGSHPx activity to 45% and phGSHPx to 29% of that in selenium-adequate controls. In iodine-deficient thyroid glands, mRNA levels were increased 2.2, 5.0 and 2.8 times for ID-I, cGSHPx and phGSHPx respectively. ID-I and cGSHPx enzyme activities were also increased but the activity of phGSHPx was decreased despite the high mRNA abundance. Thyroid selenoprotein mRNA levels were also increased in combined selenium and iodine deficiency but again there were differential effects on enzyme activities, with ID-I activity increased, cGSHPx unchanged and phGSHPx decreased. Thus, iodine deficiency may produce an oxidant stress on the thyroid gland, increasing the requirement for selenium to maintain selenoenzyme activity. When dietary supplies of selenium are limiting, thyroid selenoprotein mRNA levels are increased to compensate for overall lack of the micronutrient. Furthermore, there is a preferential supply of available selenium to ID-I and cGSHPx to allow maintenance of thyroid function.

Selenoprotein gene expression during selenium-repletion of selenium-deficient rats

Selenium repletion of selenium-deficient rats with 20 micrograms selenium / kg body weight as Na2SeO3 was used as a model to investigate the mechanisms that control the distribution of the trace element to specific selenoproteins in liver and thyroid. Cytosolic glutathione peroxidase (cGSHPx), phospholipid hydroperoxide glutathione peroxidase (PHGSHPx), and iodothyronine 5'-deiodinase (IDI) activities were all transiently increased in liver 16 to 32 h after ip injection with selenium. However, only cGSHPx and PHGSHPx activities increased in the thyroid where IDI activity was already increased by selenium deficiency. These responses were owing to synthesis of the seleoproteins on newly synthesised and/or existing mRNAs. The selenoprotein mRNAs in the thyroid gland were increased two- and threefold after the transitory increases in selenoprotein activity. In contrast, there were parallel changes in selenoprotein mRNAs and enzyme activities in the liver, with no prolonged rises in mRNA levels. The organ differences suggest that increased thryotrophin (TSH) concentrations, which are known to induce thyrodial IDI and mRNA, may control the mRNAs for all the thyroidal selenoproteins investigated and be a major mechanism for the preservation of thyroidal selenoproteins when selenium supplies are limited.

Relationship of admission thyroid function tests to outcome in critical illness

One hundred and eighty patients had serum thyrotropin, total triiodothyronine and free thyroxine concentrations measured within 3 h of admission to the Intensive Therapy Unit to assess whether thyroid function tests could predict outcome in critical illness. Overall mortality was 30.6%. Nonsurvivors were older (p = 0.001), and had higher APACHE II scores (p < 0.001) and predicted mortalities (p < 0.001). There was no difference in the median values of thyrotropin, total triiodothyronine and free thyroxine concentrations between survivors and nonsurvivors. Thyrotropin concentration was subnormal in 15 patients, normal in 152 and elevated in 13 patients. In contrast, 80 patients had subnormal triiodothyronine concentration. Free thyroxine was subnormal in five patients. Thyrotropin, total triiodothyronine and free thyroxine concentrations were not related to outcome (p = 0.360, p = 0.622, p = 0.726, respectively). No variable independently predicted death. Total triiodothyronine concentrations were lower in patients who received dopamine before admission to the intensive therapy unit than those who did not (p = 0.008); thyrotropin and free thyroxine concentrations were not influenced by dopamine administration. Serum concentrations of thyrotropin, total triiodothyronine and free thyroxine measured within 3 h of admission to the intensive therapy unit are not predictive of outcome.

The investigation of thyroid function

The measurement of thyrotropin (TSH) by a third-generation assay (functional sensitivity 0.01-0.02 mU/L) offers the single best test of assessing thyroid status and is more cost effective than measurement of thyroxine (T4) as a first-line test. However, in certain situations, e.g., nonthyroidal illness (NTI), secondary thyroid disease, early treatment of hyperthyroidism, early pregnancy, and heterophilic antibody interference, TSH alone may be misleading, and thyroid hormone measurements are also required. The author's laboratory has found that the measurement of TSH together with free thyroxine (by a nonanalogue method) offers the most speedy and effective first-line strategy for thyroid function testing. Abnormalities in thyroid function tests in euthyroid patients are most likely found in NTI and the elderly. For this reason, thyroid function tests (TFTs) should not be requested in elderly or hospitalized patients unless the presenting complaint is considered to be because of a thyroid problem.

Tissue-specific regulation of selenoenzyme gene expression during selenium deficiency in rats

Regulation of synthesis of the selenoenzymes cytosolic glutathione peroxidase (GSH-Px), phospholipid hydroperoxide glutathione peroxidase (PHGSH-Px) and type-1 iodothyronine 5'-deiodinase (5'IDI) was investigated in liver, thyroid and heart of rats fed on diets containing 0.405, 0.104 (Se-adequate), 0.052, 0.024 or 0.003 mg of Se/kg. Severe Se deficiency (0.003 mg of Se/kg) caused almost total loss of GSH-Px activity and mRNA in liver and heart. 5'IDI activity decreased by 95% in liver and its mRNA by 50%; in the thyroid, activity increased by 15% and mRNA by 95%. PHGSH-Px activity was reduced by 75% in the liver and 60% in the heart but mRNA levels were unchanged; in the thyroid, PHGSH-Px activity was unaffected by Se depletion but its mRNA increased by 52%. Thus there is differential regulation of the three mRNAs and subsequent protein synthesis within and between organs, suggesting both that mechanisms exist to channel Se for synthesis of a particular enzyme and that there is tissue-specific regulation of selenoenzyme mRNAs. During Se depletion, the levels of selenoenzyme mRNA did not necessarily parallel the changes in enzyme activity, suggesting a distinct mechanism for regulating mRNA levels. Nuclear run-off assays with isolated liver nuclei showed severe Se deficiency to have no effect on transcription of the three genes, suggesting that there is post-transcriptional control of the three selenoenzymes, probably involving regulation of mRNA stability.

Serum glutathione S-transferase B1 activity as an index of liver function in cystic fibrosis

AIMS

To evaluate serum glutathione S-transferase B1 (GST B1), a highly sensitive test of hepatocellular function, as a means of identifying liver disease in patients with cystic fibrosis (CF).

METHODS

The presence of liver disease was sought over a three year period in 60 children with CF, using a combination of clinical assessment, ultrasound examination, conventional biochemical tests of liver function (LFTs), and measurement of GST B1.

RESULTS

Reference ranges for serum GST B1 were established in a paediatric control population. The 95% value (4.55 micrograms/l) was similar to the upper limit of normal previously derived in adults. Mean (SE) serum GST B1 activities were higher in the CF population (9.0 (1.14) micrograms/l) than in age matched controls (2.4 (0.15) micrograms/l). Ten patients with CF showed clinical signs of liver dysfunction. All but one had a serum GST B1 > 4.55 micrograms/l. Twelve other patients had elevated LFTs without clinically evident liver dysfunction, six had abnormal ultrasound scans and two showed both of these anomalies. Thirty patients with CF had neither biochemical, ultrasonographic nor clinical signs of liver disease. On review three years later, clinically important liver disease was reaffirmed in eight of the 10 index cases and had become apparent in a further eight, all of whom had elevated GST B1 activities. Five (36%) of the patients with elevated LFTs and two (33%) with isolated ultrasound changes continued to show these abnormalities.

CONCLUSIONS

The limitations of conventional LFTs and ultrasound scans were evident from this study. The results suggest that elevated GST B1 activities may be a better predictor of hepatic dysfunction in CF than conventional LFTs.

Thyroidal extracellular glutathione peroxidase: a potential regulator of thyroid-hormone synthesis

Human thyrocytes were found to synthesize and secrete the selenoenzyme extracellular glutathione peroxidase (E-GPX), a process which was controlled by the Ca2+/phosphoinositol second-messenger cascade. The potential involvement of thyroidal E-GPX in the regulation of thyroid-hormone synthesis and in the protection of the thyrocyte from peroxidative damage is discussed.

Biological variation and the effect of fasting and halothane anesthesia on plasma glutathione S-transferase concentrations

Using a specific RIA, we have investigated in patients and volunteers whether fasting, diminished hepatic clearance, hemoconcentration, or within-day biological variation might be responsible for the transient increases in plasma glutathione S-transferase (GST) concentration observed after anesthesia. GST concentration was measured in 44 healthy volunteers after an overnight fast and at 3, 6, and 24 h after the fasting sample. The concentration was significantly lower at 3 and 6 h after than in the fasting sample (P = 0.0019 and P = 0.015, respectively). The change in GST concentration caused by fasting was examined in 30 subjects by comparing pre- and postfasting values. Fasting had no significant effect on GST concentration overall (P = 0.4721), but two individuals showed a marked increase in GST concentration after fasting overnight. In a separate study of 10 patients, plasma amylase activity and plasma concentrations of GST and albumin were measured immediately before and 3 h after induction of halothane anesthesia. Although GST concentration was increased at 3 h in each of the 10 patients, plasma amylase activity and plasma albumin concentration were significantly decreased in all patients (P = 0.002). Apparently, increases in GST concentration after anesthesia do not result from incidental factors.

Biological variation and the effect of fasting and halothane anesthesia on plasma glutathione S-transferase concentrations

Using a specific RIA, we have investigated in patients and volunteers whether fasting, diminished hepatic clearance, hemoconcentration, or within-day biological variation might be responsible for the transient increases in plasma glutathione S-transferase (GST) concentration observed after anesthesia. GST concentration was measured in 44 healthy volunteers after an overnight fast and at 3, 6, and 24 h after the fasting sample. The concentration was significantly lower at 3 and 6 h after than in the fasting sample (P = 0.0019 and P = 0.015, respectively). The change in GST concentration caused by fasting was examined in 30 subjects by comparing pre- and postfasting values. Fasting had no significant effect on GST concentration overall (P = 0.4721), but two individuals showed a marked increase in GST concentration after fasting overnight. In a separate study of 10 patients, plasma amylase activity and plasma concentrations of GST and albumin were measured immediately before and 3 h after induction of halothane anesthesia. Although GST concentration was increased at 3 h in each of the 10 patients, plasma amylase activity and plasma albumin concentration were significantly decreased in all patients (P = 0.002). Apparently, increases in GST concentration after anesthesia do not result from incidental factors.

Differential control of type-I iodothyronine deiodinase expression by the activation of the cyclic AMP and phosphoinositol signalling pathways in cultured human thyrocytes

The effects of TSH and the activation of the cyclic AMP (cAMP) and Ca(2+)-phosphatidylinositol (Ca(2+)-PI) cascades on the activity and expression of the selenoenzyme thyroidal type-I iodothyronine deiodinase (ID-I) have been studied using human thyrocytes grown in primary culture. Stimulation of ID-I activity and expression was obtained with TSH and an analogue of cAMP, 8-bromo-cAMP. In the presence or absence of TSH, the addition of the phorbol ester, phorbol 12-myristate 13-acetate (PMA) together with the calcium ionophore A23187, caused a decrease in ID-I activity; a decrease in ID-I expression was also observed as assessed by cell labelling with [75Se]selenite. PMA alone had no effect on ID-I activity in the presence or absence of TSH. A23187 alone produced a small but significant reduction in ID-I activity, but only in TSH-stimulated cells. These data provide evidence that the expression of thyroidal ID-I is negatively regulated by the Ca(2+)-PI cascade, and positively regulated by the cAMP cascade.

Interlaboratory/intermethod differences in functional sensitivity of immunometric assays of thyrotropin (TSH) and impact on reliability of measurement of subnormal concentrations of TSH

Clinically relevant interassay precision profiles for thyrotropin (thyroid-stimulating hormone; TSH) were constructed with human serum pools measured over 4-8 weeks by six immunometric assays, in at least two different reagent lots. Functional sensitivities (the concentration at which the interassay CV is &lt; or = 20%) were determined in four to eight clinical laboratories plus the respective manufacturer's laboratory. These studies revealed that the manufacturer's stated functional sensitivity limit is rarely duplicated in clinical practice. Loss of specificity (indicated by artifactually high values) was seen with some methods when used to measure certain unrefrigerated low-TSH sera. Measurement of TSH in four human serum pools (TSH &lt; 0.05-0.25 mIU/L) by 16 different methods (each in at least eight UK or US laboratories) showed that some methods could not reliably distinguish subnormal from normal TSH values. Better pool rankings and fewer misclassifications of low-TSH sera as "normal" were seen with use of assays capable of "third-generation" functional sensitivity (0.01-0.02 mIU/L) than with assays with "second-generation" functional sensitivity (0.1-0.2 mIU/L). Because inter- and intramethod differences in functional sensitivity negatively impact the diagnostic accuracy and cost-effectiveness of a TSH-centered thyroid-testing strategy, laboratories should independently establish an assay's functional sensitivity by a clinically relevant protocol. Moreover, manufacturers should assess functional sensitivity more realistically and improve the robustness of assays to ensure that their performance potential is consistently met in clinical practice.