Serum interleukin-6 and thyroid hormones in rheumatoid arthritis.

Using rheumatoid arthritis (RA) as a model, we have investigated whether the activation of the cytokine system, in particular, activation of interleukin (IL)-6 production, is a major cause of the depressed serum T(3) seen frequently in the nonthyroidal illness syndrome (NTIS). RA was chosen because it is a chronic autoimmune disease leading to increased serum IL-6 concentrations. We studied 16 untreated RA and 35 treated RA patients. Twenty-seven treated and 27 untreated patients with noninflammatory musculoskeletal symptoms served as controls. The patient groups displayed similar age distribution and nutritional status. Untreated RA patients displayed elevations of serum IL-6 (mean, 37.5 pg/mL) and C-reactive protein (CRP; mean, 41.3 mg/L), consistent with the inflammatory nature of their disease. Treated RA patients had significantly reduced serum IL-6 (mean, 9.9 pg/mL) and CRP (mean, 13.3 mg/L) compared with untreated RA patients, while untreated and treated patients with noninflammatory musculoskeletal symptoms had near normal serum IL-6 (mean, 2.5, 6.6 pg/mL, respectively) and CRP levels (mean, 5.8, 8.1 mg/L, respectively). However, there were no significant differences in serum concentrations of free T(3) (FT(3)) and free T(4) (FT(4)) between groups, and thyroid indices were in the normal range in RA patients. Moreover, no significant correlations between serum concentration of IL-6 and any of the thyroid hormones were demonstrated for any of the patient groups. In conclusion, we have been unable to confirm in RA that IL-6 activation leads to the low T(3) state of NTIS.

Influence of imipramine on the hypothalamic/pituitary/thyroid axis of the rat.

The effects of the tricyclic antidepressant drug imipramine at different levels of the hypothalamic/pituitary/thyroid axis were investigated in the rat. Intraperitoneal (IP) treatment for 14 days with imipramine at 10 mg/kg, but not 2 mg/kg, reduced serum total thyroxine (T4) and triiodothyronine (T3). A similar decrease in serum total T4 was observed in thyroidectomized T4-treated rats, suggesting that imipramine treatment enhances T4 clearance instead of reducing T4 secretion. There were no parallel decreases in serum free T4 and T3 concentrations, due to the simultaneous increase in the free fractions of both T4 and T3 following imipramine treatment. In vitro experiments using equilibrium dialysis indicated that neither imipramine nor its metabolite desipramine directly influenced the binding of T4 or T3 to their transport proteins following addition to normal serum, suggesting an indirect effect of imipramine or desipramine on free hormone concentrations in vivo. Concentrations of T4 and T3 in the brain, liver, and heart were unaffected by imipramine treatment, suggesting that the drug did not affect cellular uptake and metabolism of T4 and T3. Serum concentrations of thyrotropin (TSH) were unaffected by imipramine pretreatment at either dose level, compatible with the fact that serum free T4 and T3 concentrations were not reduced. Moreover, there was no difference in thyrotrope responsiveness to stimulation by TSH-releasing hormone (TRH) and to inhibition by T4 and T3 in rat anterior pituitary cells cultured ex vivo for 18 hours from control and imipramine-treated rats. Furthermore, in vitro exposure of cultured rat anterior pituitary cells to imipramine and desipramine indicated that both agents decreased TSH secretion only at concentrations greater than 10(-6) mol/L. These concentrations of imipramine and desipramine in the culture medium would exceed the free concentrations of these drugs seen in vivo therapeutically. In addition, no direct effects of 10(-6) mol/L imipramine or desipramine on the TSH response to TRH or to T3 were observed in vitro in cultured pituitary cells. A potential indirect effect of imipramine or desipramine on TSH secretion via altered hypothalamic control of thyrotropes does not seem likely, due to the lack of effect of imipramine treatment on serum TSH concentrations in imipramine-treated rats. In conclusion, imipramine treatment reduces serum total T4 and T3 in the rat, with enhanced clearance being the most likely explanation for the effect on T4. There was no evidence for altered tissue T4 or T3 concentrations or for altered thyrotrope function. The enhanced T4 clearance may explain the reduction in total T4 reported for imipramine-treated depressed patients. However, the effects of imipramine treatment on the transport of thyroid hormones in plasma need to be examined in more detail in patients, since interspecies differences in the nature of the transport proteins preclude extrapolation of the present results from the rat.

Effect of interleukin-1 beta, tumour necrosis factor-alpha and interleukin-6 on the control of thyrotropin secretion.

Since serum concentrations of tumour necrosis factor-alpha (TNF-alpha), interleukin-1 beta (IL-1 beta), and interleukin-6 (IL-6) are elevated in infectious and inflammatory illnesses, we examined their potential role in contributing to the low TSH concentrations associated with such conditions, both at the level of the pituitary and the hypothalamus. 20 hours exposure to recombinant murine TNF-alpha (10(-11) to 10(-10) mol/l) enhanced the basal and the TRH-stimulated release of TSH by cultured rat anterior pituitary cells, but 4 hours exposure increased only basal TSH secretion. Recombinant human (rh) IL-1 beta, at a dose of 10(-11) mol/l only, produced a very small increase in basal TSH secretion after 4h, but not 20h, exposure. TRH-stimulated TSH secretion was not affected by IL-1 beta in concentrations up to 10(-10) mol/l, at either exposure time. Rh IL-6 (10(-12) to 10(-9) mol/l), had no effect on basal or TRH-stimulated TSH secretion at either exposure time. TNF-alpha, IL-1 beta, and IL-6 all failed to modify the inhibitory response to triiodothyronine (T3) and thyroxine (T4) on TSH secretion, under basal or TRH-stimulated conditions. Indirect effects of the cytokines on the stimulation or inhibition of TSH secretion, via TRH or SRIF respectively, were tested in isolated rat hypothalamic slices. 30 min exposure to TNF-alpha, IL-1 beta, or IL-6 had no effect on the basal release of SRIF. However, IL-1 beta, from 2.5 x 10(-12) to 10(-10) mol/l, produced a dose-dependent enhancement of the SRIF released by 5 x 10(-2) mol/l extracellular K+. The effect appeared to be mediated via IL-1 receptors, and to involve prostanoid formation, since it was inhibited by IL-1 receptor antagonist protein, 10(-7) mol/l, and indomethacin, 2.8 x 10(-5) mol/l, respectively. Neither basal nor K(+)-stimulated TRH release was influenced by TNF-alpha, IL-1 beta, or IL-6. The results indicate that direct effects of these cytokines on the pituitary do not contribute to reduced circulating TSH concentrations during inflammation and infection, but that enhanced hypothalamic release of SRIF, in response to elevated IL-1 beta, could contribute to such a decrease in TSH. None of the cytokines tested decreased hypothalamic TRH release in vitro. However, further in vivo experiments would be required to determine whether a longer exposure to these agents could reduce TRH release either directly, or indirectly via inputs from outside the hypothalamus.

The effect of oleic acid on the secretion of thyrotrophin and growth hormone by cultured rat anterior pituitary cells.

Elevation of non-esterified fatty acids (NEFA) in vivo is associated with abnormal control of TSH. To determine whether TSH secretion is directly inhibited by NEFA, as has been reported for GH, cultured rat anterior pituitary cells were exposed for 20 h to oleic acid in medium containing 7.7 x 10(-5) mol/l bovine serum albumin (BSA). In a molar ratio with albumin of 1.2 (total oleic acid 9 x 10(-5) mol/l), or greater, oleic acid inhibited basal GH secretion (maximum inhibition to 40% of control) while basal TSH was less affected, a ratio of 3 (2.3 x 10(-4) mol/l oleic acid) or greater causing a smaller degree of inhibition (maximum inhibition to 80% of control). In the presence of 10(-9) mol/l growth hormone-releasing hormone or 10(-8) mol/l TRH, inhibition was achieved at a ratio of 12 (9 x 10(-4) mol/l oleic acid) or greater. Basal TSH was less sensitive to inhibition by thyroxine (T4) in the presence of oleic acid/albumin at a ratio of 6 or greater, and inhibition by oleic acid was less than additive with T4 at a ratio of 6 or greater. Responses to tri-iodothyronine (T3) were unaffected at a ratio of 6 (4.6 x 10(-4) mol/l oleic acid), but a ratio of 12 inhibited the effects of both T3 and T4 on TSH.(ABSTRACT TRUNCATED AT 250 WORDS)

Endocrine and cytokine changes during elective surgery.

Elective surgery was used as a model of severe non-thyroidal illness (SNTI) to study the inter-relation between changes in serum thyroid hormones, thyroid stimulating hormone (TSH), cortisol, and interleukin 6 concentrations. The study was designed to determine whether the expected interleukin 6 increases after surgery are the cause of decreased serum tri-iodothyronine (T3) concentration normally observed following severe trauma. Blood was sampled for 24 hours before, during, and for 48 hours after abdominal surgery under general anaesthesia in 11 patients. Total T3 decreased 30 minutes after induction and continued to decrease at 24 hours. After a transient increase at 30 minutes, free T3 also decreased, and free thyroxine (T4) concentrations, other than a similar transient increase, did not change. TSH concentrations were increased at four hours and the nocturnal surge was suppressed. The increase in the serum interleukin 6 concentration was not observed until four hours. Cortisol concentrations were increased at 30 minutes and peaked at four hours. Therefore, the early changes in thyroid hormones and TSH accompanying surgery do not seem to be caused by changes in interleukin 6 concentrations.

Effect of fatty acids on rat liver nuclear T3-receptor binding.

The effects of selected fatty acids (linoleic, oleic, and palmitic) on triiodothyronine (T3)-receptor binding were compared in isolated rat hepatocytes, rat liver nuclei, and receptor protein. Scatchard analysis indicated that the inhibition of T3-receptor binding by fatty acids was characterized by an increase in Kd and no change in maximum binding capacity (MBC). In isolated receptors, the rank order of potency for inhibition was linoleic acid greater than oleic acid greater than palmitic acid. The Ki for oleic acid in isolated receptors was the same as that for whole nuclei (15.4 +/- 1.3 v 16.3 +/- 1.9 mumol/L, respectively), indicating that the inhibition of nuclear T3 binding is probably at the level of the receptor protein itself. In isolated hepatocytes, linoleic acid was more potent than oleic acid in inhibiting T3 binding to nuclear receptors. Cell-associated T3 was not affected by the presence of fatty acids, implying that cellular uptake of T3 was not inhibited. High concentrations of fatty acids were necessary for inhibition of T3-receptor binding in isolated hepatocytes, with linoleic acid being one to two orders of magnitude less potent in isolated hepatocytes compared with isolated receptors (Ki, 179 +/- 12 v 4.4 +/- 0.5 mumol/L, respectively). It is concluded that the inhibitory effect of fatty acids on T3-receptor binding in isolated rat hepatocytes probably occurs at the level of the nuclear receptor, and does not involve an inhibition of the access of T3 to the receptor. However, in vivo it seems unlikely that fatty acids will have access to the nuclear receptors in sufficiently high concentrations to affect T3-receptor binding in liver cells.

Critical assessment of the Amerlite free T4 assay.

The measurement of serum free T4 (FT4) by analogue methods has been severely criticised because the T4 analogue binds to albumin. Amersham have recently introduced a method utilising horseradish peroxidase-labelled-T4 (HRP-T4) designed to overcome this problem and have incorporated it into the Amerlite enhanced luminescence immunoassay system. We have critically evaluated this method for its analytical and clinical validity. Experiments in which anti-albumin was added to normal serum suggested that the HRP-T4 label did not bind to endogenous albumin while the addition of albumin caused no significant change in FT4 concentration. Adding oleic acid up to 5 mmol/L to simulate increased non-esterified fatty acid concentration did not increase the apparent FT4. Serum sampled from subjects independently allocated to clinical groups were compared with an euthyroid group. The untreated hyperthyroid group values were distinctly elevated while the untreated hypothyroid group were appropriately low. Oestrogen therapy, low TBG, familial dysalbuminaemic hyperthyroxinaemia and non-thyroidal illness groups all reflected their euthyroid status, as did pregnancy samples which also showed a tendency to lower values in late pregnancy, consistent with previous observations. In conclusion, the Amerlite FT4 method appears to overcome some of the problems associated with analogue methods. A small survey showed it to be diagnostically valid in a wide variety of clinical states.

Do nonesterified fatty acids displace thyroxin from its plasma binding sites in severe nonthyroidal illnesses?

Severe nonthyroidal illnesses have been associated with increases in nonesterified fatty acids (NEFA) and the dialyzable fraction of thyroxin (T4) in plasma. We have further investigated their possible relationship in severe nonthyroidal illnesses as well as in induced in vivo and in vitro situations involving increased NEFA. We demonstrate that there is no relationship between NEFA and the dialyzable fraction of T4, either in severe nonthyroidal illnesses or in the other situations, unless plasma NEFA concentrations exceed 5 mmol/L in normal persons or 1.7 mmol/L in nonthyroidal illnesses, and that this concentration was not reached in the patients we studied, with one exception. We conclude that NEFA are unlikely to contribute to an inhibition of the binding of T4 to the binding proteins that might be present in plasma of patients with severe nonthyroidal illnesses unless their NEFA concentrations are very high.

Familial dysalbuminaemic hyperthyroxinaemia and other causes of euthyroid hyperthyroxinaemia.

Clinical and biochemical studies on a family in which 3 members have familial dysalbuminaemic hyperthyroxinaemia (FDH) are presented. They were clinically euthyroid with elevated serum thyroxine (T4) and free T4 indices but normal free T4 by equilibrium dialysis and normal serum triiodothyronine (total and free). All thyroid function tests on the remaining family members were normal. The inheritance is consistent with autosomal dominance. Also presented are data on 4 unrelated patients with FDH and two patients with T4 autoantibodies. The methods for detecting FDH, T4 antibodies and other causes of euthyroid hyperthyroxinaemia are now freely available. Since these anomalies may be more common than previously supposed, clinical awareness of the conditions is necessary to protect patients from the consequences of incorrect diagnosis of thyrotoxicosis.

Importance of postmortem changes in measurements of thyroid function in studies of sudden infant death syndrome.

Various measurements of thyroid function were made from serum sampled at necropsy by cardiac puncture in adults and compared with the equivalent values before death. There were significant increases in serum total T3 (0.66 before death, 1.14 nmol/l after death), free T3 (0.82 to 1.52 pmol/l), and reverse T3 (1.12 to 1.75 nmol/l); and significant decreases in total T4 (73.7 to 56.6 nmol/l), free thyroxine index (FTI) (83.2 to 56.6), and free T4 (14.2 to 11.9 pmol/l). The increase in serum T3, presumably from T4 deiodination after death, observed in this study may explain the increased T3 concentration reported in serum sampled from cases of sudden infant death syndrome (SIDS).

A rat model of the 131I-induced changes in thyroid function.

Thyrotoxic patients treated with Iodine-131 (131I) often present with low thyroxine (T4), normal triiodothyronine (T3) and raised thyrotropin (TSH) concentrations in serum. We have developed a rat model of this low T4, raised TSH state. Rats were injected with 50, 150 or 450 mu Ci 131I. A dose of 50 mu Ci 131I caused no significant effect on thyroid function, as assessed by serum parameters whereas both 150 mu Ci and 450 mu Ci 131I caused a significant fall in serum T4 concentration accompanied by a significant rise in TSH concentration. In all groups serum T3 concentration was not significantly altered when compared to controls. The clearance of 131I from the rats showed a single exponential curve (t 1/2 3.38 +/- 0.61 days) over the range of 131I doses used. Differing body weights had no effect on the serum T4 changes induced by 131I.

Studies on the accuracy of a quantitative kit radioimmunoassay for free thyroxine.

Experiments designed to determine the accuracy of the Corning Immophase Free T4 assay revealed that there was a marked perturbation of thyroxine binding to the serum proteins during the assay; binding of 125I-T4 to the immobilised T4 antibody in the absence of merthiolate increased considerably with buffer dilution of serum; and the free thyroxine concentration declined significantly during pregnancy. These changes in measured fT4 were not observed with an equilibrium dialysis-radioimmunoassay procedure. The assay was precise, easily performed, and as effective as the free thyroxine index (FTI) in diagnosing thyroid disease. We conclude that the assay does not provide an accurate quantitative estimation of serum fT4 concentration in samples with elevated TBG concentration and that the kinetic principles on which the assay is based are altered as serum binding-protein concentration is reduced (serum dilution). Both FTI and fT4 (Corning) data in pregnant patients should be interpreted with caution and with reference to clinical symptomatology and other thyroid function tests.

Slow thyroxine binding globulin in the pathogenesis of increased dialysable fraction of thyroxine in nonthyroidal illnesses.

The expected decreases of serum total T4, total T3, T4 binding prealbumin, and T4 binding globulin (TBG) concentrations were found in a selected series of patients with nonthyroidal illnesses (NTI). An increase in percent dialysable fraction of T4 was also found. Although serum TBG concentrations were decreased, the proportion of a slow moving form of TBG, designated slow TBG (STBG) was increased, the absolute concentration being not significantly different from that measured in controls. Thus the observed decrease in TBG in NTI occurs in the normal TBG fraction which binds T4 much more avidly than does STBG. It is suggested that the increase in the proportion of STBG, due to a decrease in normal TBG, has an important role in the pathogenesis of the increase in percent dialysable fraction of T4 in serum from patients with NTI.

Retarded fetal brain development resulting from severe dietary iodine deficiency in sheep.

Sheep have been used to study the effect of dietary iodine deficiency on the development of the fetal brain. Severe iodine deficiency caused reduction in fetal brain and body weights and in brain DNA and protein from 70 days gestation to parturition. The lowered brain weight and brain DNA at 70 days gestation indicates a reduced number of cells, probably due to slower neuroblast multiplication which normally occurs from 40-80 days in the sheep, and the reduction in DNA and protein after 80 days implies that the development of neuroglia could be slowed also in iodine deficiency. Morphological changes were observed in both the cerebral hemispheres and the cerebellum. In the cerebral hemispheres of the iodine-deficient fetuses an increased density of neurons was apparent histologically in the motor cortex and visual cortex and in the CA1 and CA4 areas of the hippocampus in comparison with controls. In the cerebellum there was delayed migration of cells from the external granular layer to the internal granular layer and increased density of Purkinje cells in the iodine-deficient fetal brains. In addition, the molecular area was increased and the medullary area reduced in comparison with controls. These change are indicative of delayed brain maturation. Evidence of fetal hypothyroidism was provided by low fetal thyroid iodine and plasma T4 values, thyroid hyperplasia from 70 days gestation, significant reduction in body weight at the same time as the brain retardation, and absence of wool growth and delayed skeletal maturation near parturition. It is apparent from the biochemical and histological changes observed during iodine deficiency that iodine is an essential element for normal fetal brain and physical development in the sheep.

Elevated serum thyroxine-binding globulin by X-chromosome transmission.

A kindred with a high prevalence of hereditary serum thyroxine-binding globulin (TBG) is described. Seventeen of th 29 members of the kindred have the increased TBG trait as demonstrated by a combination of increased serum total thyroxine (T4) and total triiodothyronine (T3) and decreased T3 resin uptake. In 12 of the 17, the anomaly was confirmed by measuring serum TBG activity as maximum binding capacity of T4. The pattern of increased TBG is consistent with X-linked inheritance. One of the affected members had proven thyrotoxicosis and two others were subjected to sub-total thyroidectomy.

Thyroid-stimulating hormone (TSH) receptor antibodies and antibodies stimulating adenyl cyclase in relatives from two families with a high prevalence of Graves' hyperthyroidism: a ten-year follow-up study.

Thyroid-stimulating hormone (TSH) receptor antibodies and antibodies stimulating adenyl cyclase were measured in 47 relatives of patients with Graves' hyperthyroidism from two families with a high prevalence of the disease, in whom bioassays for the long-acting thyroid stimulator (LATS) had been performed 10 years earlier. Tests were also carried out in six propositi from the two families and age- and sex-matched normal subjects from six families. There had been no new cases of hyperthyroidism since the first study, although one subject was clinically and biochemically hyperthyroid at the time of study and two more were biochemically borderline hyperthyroid but clinically euthyroid. Levels of serum T4 thyrotropin, and percentage T3 resin uptake and free thyroxine indices were similar for relatives and normal subjects, although the mean serum T3 level for relatives was significantly greater than that for the normal subjects. Antibodies were not detected by either assay in any relative. Significant titers of antithyroglobulin antibodies were demonstrated in 4 of 44 relatives but in none of 46 normals tested, while thyroid cytoplasmic antibodies were detected in 8 of 44 relatives and 3 of 45 normals. The mean serum IgG for Graves' relatives was significantly greater than that for the normals, although the mean IgM and IgA levels for the two groups were not significantly different.

Evaluation of a new free-thyroxin assay.

The Amerlex Free Thyroxin (T4) Radioimmunoassay Kit (Amersham International Ltd.) is a new direct equilibrium radioimmunoassay for free T4 based on an antiserum with very high affinity for T4, and a unique 125I-labeled T4 analog as tracer. It is a very simple single-tube radioimmunoassay, making use of Amerlex particles to separate antibody-bound from free species. Interassay precision (CV) is 3.7% at 13 pmol/L and 2.3% at 30 pmol/L; within-assay precision is 4.2% at 21 pmol/L. The reference interval is 11-22 pmol/L. The assay did not misclassify any patients tested who had untreated myxedema or untreated thyrotoxicosis. The free T4 assay excelled both the free T4 index and the T4/T4-binding globulin ratio in correcting for increased thyroxin-binding globulin from pregnancy, and it was better than the index but not better than the ratio in correcting for increased thyroxin-binding globulin in users of oral contraceptives.