The effect of thyrotrophin-releasing hormone on plasma prolactin and thyrotrophin levels in primary hypothyroidism

Medical causes and consequences of hyperprolactinaemia. A context for psychiatrists

Hyperprolactinaemia is the commonest endocrine disorder of the hypothalamic-pituitary axis and can lead to both short-term sexual dysfunction and galactorrhoea, and long-term loss of bone mineral density. Prolactin is secreted from the anterior pituitary gland under the influence of dopamine, which exerts a tonic inhibitory effect on prolactin secretion. Physiological regulators of prolactin secretion include many different types of 'stress' and sleep. Disruption of the normal control of prolactin secretion results in hyperprolactinaemia from pathological and pharmacological causes. The administration of antipsychotic medication is responsible for the high prevalence of hyperprolactinaemia in people with severe mental illness. Physiological hyperprolactinaemia, such as pregnancy and lactation, should be distinguished from pathological causes to prevent unnecessary investigation and treatment. The causes, consequences and management of hyperprolactinaemia are discussed in this article.

Contemporary management of prolactinomas

Prolactin-secreting pituitary adenomas--prolactinomas--are the most common type of functional pituitary tumor. Treatment of hyperprolactinemia is indicated because of the consequences of infertility, gonadal dysfunction, and osteoporosis. Making the correct diagnosis is important because the first line of therapy is medical management with dopamine agonists. Medical therapy is effective in normalizing prolactin levels in more than 90% of patients, but longterm treatment may be required in some patients. Transsphenoidal surgery is usually indicated in those patients in whom medical therapy fails or cannot be tolerated, or in patients who harbor microprolactinomas. In experienced hands, a hormonal and oncological cure can be achieved in more than 90% of patients after transsphenoidal removal of microprolactinomas with minimal risks. Thus, surgery may be an option for microprolactinomas in a young patient who desires restoration of fertility and avoidance of long-term medical therapy. The authors review the diagnosis and management of prolactinomas, including medical therapy, surgical therapy, and stereotactic radiosurgery.

The prevalence of hyperprolactinaemia and association with markers of autoimmune thyroid disease in survivors of the Whickham Survey cohort

OBJECTIVE

Few data exist on the prevalence of hyperprolactinaemia in the community. This study was intended to determine the prevalence of hyperprolactinaemia in a sample closely matched to the current British population aged 38 years and over.

DESIGN AND PATIENTS

The 1877 survivors at the 20-year follow-up of the Whickham Survey were a cross-sectional sample of the community aged 38 years and over. Serum was frozen and stored at -30 degrees C from 90% of the survivors (751 men, 924 women, median age 58 years (range 38 to 93 years)) who participated in the follow-up survey.

MEASUREMENTS

Two years after the follow-up survey, serum prolactin concentrations were measured by ELISA/1 step sandwich assay (reference range < or = 600 mU/l in men and women). A repeat prolactin measurement was made in those subjects who had prolactin levels within the top 2.5% of men and women in this sample.

RESULTS

At screening, 0.7% of the men and 2.5% of the women had serum prolactin levels greater than 600 mU/l. For men, 2.5% were above 400 mU/l. The prevalence of hyperprolactinaemia, if defined as greater than 400 mU/l in men and greater than 600 mU/l in women on repeat testing, was 1.4% in the men and 1.2% in the women. The aetiology in men was prolactin-raising drugs (n = 3), renal failure (n = 1), microprolactinoma (n = 1), and unknown (n = 2), and in women it was prolactin-raising drugs (n = 7), microprolactinoma (n = 1), and unknown (n = 1). Logarithmic transformation of serum prolactin concentrations produced Gaussian distributions with 95% reference ranges of 60-430 mU/l in men and 40-560 mU/l in women. No significant relationship was found in either sex between hyperprolactinaemia and age or evidence of autoimmune thyroid disease at either survey. In women, there was no association with age, distance beyond the menopause or duration of reproductive years but prolactin levels were slightly higher in those on oestrogen therapy (geometric mean prolactin 226 mU/l compared to 178 mU/l; t-test on log prolactin t = 3.79; P < 0.0001).

CONCLUSIONS

This study has demonstrated that a gender-related reference range for serum prolactin is necessary. Pituitary pathology is not common and screening with measurement of serum prolactin is not warranted in middle-aged and elderly subjects. In asymptomatic subjects with modestly elevated serum prolactin levels (< 3 SD above the mean), extensive pituitary imaging and investigation is unwarranted. Autoimmune thyroid disease was not a significant cause of hyperprolactinaemia in this sample.

Empty sella developing during thyroxine therapy in a patient with primary hypothyroidism and hyperprolactinaemia

A 35 year old woman presented with severe primary hypothyroidism and galactorrhea. A very high prolactin level was also detected and computerized tomography scan of the sellar region demonstrated an enlarged pituitary gland associated with contrast enhancement. Replacement therapy with thyroxine corrected both biochemical and clinical abnormalities but empty sella developed during this therapy. It is concluded that empty sella may be related to thyroxine-induced shrinkage of lactotroph and/or thyrotroph cell hyperplasia.

Effects of protracted hypothyroidism on pituitary function and structure in endemic cretinism

Pituitary function and structure were assessed in 69 endemic cretins from western China. In hypothyroid cretins (TSH greater than 10 mIU/l), CT imaging of the pituitary revealed adenoma in five of 20 (25%) and partially empty sella (PES) in a further eight of 20 (40%). The majority of tumours were microadenomas and showed a relation with higher levels of serum TSH but not with duration of hypothyroidism. Dynamic pituitary testing with TRH and GnRH in four patients with adenoma on CT gave a flat TSH response but significant rises in serum PRL, GH, LH and FSH concentrations. Hyperprolactinaemia (greater than 350 mIU/l) was present in hypothyroid cretins only (13 of 26; 50%) and serum PRL showed a curvilinear relation with serum TSH levels (r = 0.7, P less than 0.0001). Hypogonadism was seen in approximately half the cretins with high PRL levels. Our data suggest that severe protracted thyroid hormone deficiency may result in thyrotrophin adenomas of the pituitary gland. Disturbances of growth, puberty, and sexual function in endemic cretins are explained by the secondary effects of thyroid hormone deficiency on pituitary function.

Prolactin in patients with major depressive disorder and in healthy subjects. I. Cross-sectional study of basal and post-TRH and postdexamethasone prolactin levels

Prolactin (PRL) levels were investigated in patients with major depressive illness and in healthy subjects. Basal and postdexamethasone levels were measured in 27 patients, and levels after thyrotropin-releasing hormone (TRH) stimulation (delta PRL) were measured in 22 patients. Basal and delta PRL were also determined in 64 age- and sex-matched healthy subjects. Both basal and postdexamethasone PRL levels were normal in depressed patients, with the postdexamethasone levels in particular showing no correlation to postdexamethasone cortisol concentrations. One milligram oral dexamethasone did not influence 4:00 PM PRL levels in 15 healthy subjects. delta PRL was significantly elevated in both male and female patients. These increases were not correlated with severity of illness and are difficult to interpret owing to the complexity of the PRL regulatory system. No significant correlations were found between basal or post-TRH PRL and cortisol, thyroid-stimulating hormone (TSH), thyroid hormones, gonadotropins, or estradiol in the patients. However, surprisingly significant positive correlations between basal PRL and basal cortisol, T4 and reverse T3 occurred in healthy subjects. It is as yet unclear how this finding can be explained and what relevance it has. Women tended to have higher basal PRL concentrations than men, but the difference was not significant in either group. delta PRL was significantly higher in women than in men in both patients and controls. No significant influence of age was found.

Prolactin in patients with major depressive disorder and in healthy subjects. II. Longitudinal study of basal prolactin and post-TRH-stimulated prolactin levels

Longitudinal investigations of basal prolactin (PRL) and prolactin concentrations following thyrotopin-releasing hormone (TRH) stimulation (delta PRL) were conducted in 17 patients with major depressive disorder and healthy subjects. The patients were being treated with either clomipramine or maprotiline. Both basal and delta PRL increased significantly after clinical response during treatment with both drugs. However, these increases in basal and delta PRL were independent of each other. Surprisingly, elevations of basal PRL were significantly greater in responders than in nonresponders, whereas those in delta PRL showed no corresponding significant difference. These results suggest that the two drugs stimulate basal and delta PRL by different mechanisms. The increases in basal prolactin levels found in responders may possibly be due to weaker inhibition of prolactin due to "down-regulated" beta adrenergic receptors and/or enhanced activity of supersensitive serotonergic receptors. Neither basal PRL nor delta PRL proved to be a predictor of therapy response. The intraindividual retest reliabilities of both basal and delta PRL in healthy subjects was so good that a single blood sample would seem to be sufficient for investigating most issues involving PRL in psychiatric patients.

Hyperprolactinaemia and microadenomas in primary hypothyroidism

The presence at presentation of hyperprolactinaemia and abnormalities on high resolution computed tomographic (CT) scanning of the pituitary gland have been documented in 19 patients with primary hypothyroidism. The changes on thyroid hormone replacement were followed for up to 28 months after presentation. At presentation serum PRL was elevated in 12 subjects (454 to 2612 mU/l); after stabilization of thyroid replacement PRL remained raised in 9 patients (525 to 1888 mU/l) despite 10-18 months of treatment. Pretreatment CT scans showed enlarged pituitaries in 11 patients with suprasellar extension in five. In 11 patients there was initially an area of low attenuation suggesting a microadenoma and on rescanning this appearance remained in five patients. Hyperprolactinaemia in hypothyroid patients does not always return to normal with thyroid hormone replacement.

Diminution of prolactin and increase of thyrotrophin response to thyrotrophin releasing hormone occurring independently of changes in basal level of these hormones during treatment of hypothyroidism

In order to study the effect of rising thyroid hormone levels in hypothyroidism upon prolactin responses to TRH, in relation to changes in basal prolactin concentration and to changes in TSH responses, we followed 18 patients by performing TRH tests before and after 14, 28, 42 and 56 d on gradually increasing levothyroxine dosages, plus a final test when euthyroidism was achieved. Basal prolactin rose initially, at day 14, followed by a return on days 28, 42 and 56 to a level similar to the pretherapeutic value, while at euthyroidism prolactin had fallen below the original value. The area under the prolactin response curve on TRH stimulation was unchanged at 14 d, notwithstanding the rise of basal prolactin concentration. At 28, 42 and 56 d the responses declined progressively along with basal prolactin concentrations that remained equivalent to the pretreatment level. The TSH response to TRH rose at 14, 28, and 42 d along with a continuous downward course of basal TSH. Thus, substitution with L-thyroxine inhibited the responsivity of prolactin but enhanced that of TSH to TRH. The fact that these changes occurred in opposite directions, appears to rule out a negative feed-back effect of T4 on hypothalamic TRH secretion. The effects on responses were shown to occur independently of those on basal secretion of prolactin and TSH.

Prolactin secretion and nodular goiter. (Hypothetical correlations)

In a group of 129 female patients with nodular goiter, Graves' disease and primary hypothyroidism, the PRL and TSH secretions were studied in parallel. Concurrent changes in the two hormones were evident only in conditions of marked hyper- or hypothyroidism. In the patients with non-toxic nodular (single or multiple) goiter the TSH showed low or normal values, while the PRL appeared normal or elevated. From these results two important conclusions can be drawn: 1. the T3 and T4 levels interact with PRL secretion concomitantly with TSH only when they undergo a huge deviation from the normal range; 2. the goitrogenic action of PRL that has been reported in experimental animals cannot be excluded in man.

Dissociation between TSH and prolactin dynamics in treated thyrotoxicosis

TSH and prolactin secretory patterns in thyroid disease have generally been reported as concordant. We studied TSH and prolactin responses to TRH infusion (500 mug) in euthyroid individuals previously treated for thyrotoxicosis with 131I or antithyroid drugs. The 131I-treated group (seven men, twenty women) had been clinically and biochemically euthyroid (normal serum thyroxine and triiodothyronine levels) for 6 months to 4-5 years (kappa +/- SD = 17-1 +/- 4-1 months). Based on maximal TSH increment (deltaTSH), three patient groups were identified: Group 1 [normal deltaTSH, n = 6]: delta prolactin was normal in two, blunted in one and exaggerated in three. Group 2 (exaggerated TSH response, n = 8): delta prolactin was normal in two, blunted in one and exaggerated in five. Group 3 (TSH nonresponders, n = 13): delta prolactin was normal in five, blunted in three, and exaggerated in five. Eleven patients (three men, eight women) were studied after 6 months antithyroid-drug treatment. All were clinically and biochemically euthyroid. All but one showed a blunted TRH-TSH response. All three men showed an exaggerated delta prolactin as did four of eight women. Three women showed a blunted delta prolactin and in one, delta prolactin was normal. Thus, TRH-induced TSH and prolactin response patterns in treated thyrotoxicosis are not uniformly concordant, and, while a blunted or absent TSH response commonly persists long after euthyroidism has been restored, this is most frequently accompanied by a normal or exaggerated prolactin response.

The clinical utility of a heterologous radioimmunoassay for human prolactin

A specific and sensitive double antibody radioimmunoassay for human prolactin has been developed utilizing anti-human prolactin and purified ovine prolactin for radioiodination. Employing this radioimmunoassay, as little as 1.5 ng of prolactin can be quantified in a ml of human serum. The prolactin in serum detected by the radioimmunoassay behaved in a similar manner to purified human 125-I-labled prolactin on polyacrylamide gel electrophoresis. Exogenous prolactin could be completely recovered when varying amounts were added to a constant volume of serum. In normal control and primary hypothyroid subjects, intravenous administration of thyrotropin-releasing hormone elicited a peak prolactin response 30 min post-injection. Hyperthyroid and hypopituitary subjects did not manifest a significant increase in serum prolactin concentration after the injection. Subjects with functional galactorrhea had high resting prolactin levels and exhibited peak increase in serum prolactin concentration at 60 min following the administration of thyrotropin-releasing hormone.