Alterations in thyroid hormones and thyrotropin response to TRH in cirrhotic patients with or without hyperammonemia (author's transl)]

Serum concentrations of thyroid hormones and TSH response to TRH are studied in normal controls and in patients with liver cirrhosis. T3 levels are significantly diminished in the cirrhotic group while the mean total T4 concentration, the basal TSH and the magnitude of the peak response to TRH are similar in cirrhotic and control patients. However when the cirrhotic patients are separated in two groups according to the level of arterial ammonemia, it appears that the group with hyperammonemia (n = 10) has a significantly higher peak response than the control group. Since hyperammonemia is a wittness of hepatic encephalopathy it is suggested that TSH release or synthesis may be modified by this situation resulting from a cerebral accumulation of false neurotransmitters and a depletion of aminergic mediators.     

GROWTH HORMONE HYPERRESPONSIVENESS TO GROWTH HORMONE-RELEASING HORMONE IN PATIENTS WITH SEVERE LIVER CIRRHOSIS

Patients with severe liver disease often have high baseline plasma GH levels and/or paradoxical GH release in response to several secretagogues, e.g. TRH. In this paper, we have investigated in a group of cirrhotic patients the GH response to GH-releasing hormone (GHRH) and evaluated the effectiveness of GHRH to cause GH release in TRH responder and non-responder patients. Ten patients and seven age- and sex-matched control subjects were given bolus injections of GHRH (0.1 and 1.0 microgram/kg i.v. on separate occasions). GHRH 0.1 microgram/kg failed to cause a GH response in both control and cirrhotic subjects, but 1.0 microgram/kg caused a significantly higher GH response in patients than in controls. Evaluation of the GH response curve after GHRH revealed a similar pattern of secretion in the TRH-responders (four subjects) and non-responders (six subjects). These results suggest that the enhanced GH responsiveness to GHRH in cirrhotic patients may contribute to their high baseline GH levels and/or secretory rate, and the mechanism(s) of the paradoxical GH rise after TRH seems to be separate from that for GH hyperresponsiveness to GHRH.     

Plasma growth hormone profile in acromegaly before and ten days after transsphenoidal surgery

Profiles of plasma GH, plasma somatomedin-C and serum PRL concentrations as well as serum GH response to iv TRH were determined in 11 patients with acromegaly before and 10 days after surgery. Blood for profile determinations was drawn from a peripheral vein with a continuous withdrawal pump changing the recipient tube at 30-min intervals. Before surgery all patients had high plasma GH concentrations with irregular peaks and somatomedin-C concentrations were elevated. The response to TRH was abnormal in 8 patients. Three patients had slightly elevated PRL concentrations and one had high PRL concentration (6900 mU/l). Ten days after surgery GH concentrations were still high in 2 patients (greater than 5 mU/l), as were somatomedin-C concentrations (3.2 and 2.4 U/l, respectively). In 3 patients basal GH concentrations were less than 5 mU/l and somatomedin-C concentrations were normal, but there were no major peaks in plasma GH concentrations. In 2 patients major peaks in GH concentrations appeared after surgery, but basal GH concentrations were 1.9 and 0.95 mU/l, respectively. One patient with hyperprolactinemia still had slightly elevated PRL concentration (486 mU/l), but the response to TRH was normalized. Finally, in 4 patients, mean GH concentrations were markedly reduced, somatomedin-C concentrations normalized and apparently normal plasma GH profiles appeared with low or undetectable basal levels separating major peaks. The results indicate that in some patients with acromegaly apparently normal GH secretion can be demonstrated 10 days postoperatively. Characterization of circadian GH rhythms during the early postoperative stage may contribute to the evaluation of the effect of surgery.     

Decreased growth hormone (GH) response to oral clonidine in endemic cretinism: effect of L-T3 therapy

As GH secretion is dependent upon thyroid hormone availability, the GH responses to clonidine (150 micrograms/m2) and the TSH and PRL response to TRH were studied in eight endemic (EC) cretins (3 hypothyroid, 5 with a low thyroid reserve) before and after 4 days of 100 micrograms of L-T3. Five normal controls (N) were also treated in similar conditions. Both groups presented a marked increase in serum T3 after therapy (N = 515 +/- 89 ng/dl; EC = 647 +/- 149 ng/dl) followed by a decrease in basal and peak TSH response to TRH. However, in the EC patients an increase in serum T4 levels and in basal PRL and peak PRL response to TRH after L-T3 therapy was observed. One hypothyroid EC had a markedly elevated PRL peak response to TRH (330 ng/dl). There were no significant changes in basal or peak GH values to treatment with L-T3 in normal subjects. In the EC group the mean basal plasma GH (2.3 +/- 1.9 ng/ml) significantly rose to 8.8 +/- 3.2 ng/ml and the mean peak response to clonidine (12.7 +/- 7.7 ng/ml) increased to 36.9 +/- 3.1 ng/ml after L-T3. Plasma SM-C levels significantly increased in N from 1.79 +/- 0.50 U/ml to 2.42 +/- 0.40 U/ml after L-T3 (p less than 0.01) and this latter value was significantly higher (p less than 0.05) than mean Sm-C levels attained after L-T3 in the EC group (respectively: 1.14 +/- 0.59 and 1.78 +/- 0.68 U/ml). These data indicate that in EC the impaired GH response to a central nervous system mediated stimulus, the relatively low plasma Sm-C concentrations, and the presence of clinical or subclinical hypothyroidism may contribute to the severity of growth retardation present in this syndrome.     

DISTINCTIVE FEATURES OF PROLACTIN SECRETION IN ACROMEGALIC PATIENTS WITH HYPERPROLACTINAEMIA

We have investigated the relationship between the plasma PRL concentrations of 98 untreated acromegalic patients and the GH levels during basal and dynamic conditions. Hyperprolactinaemia was present in 27 patients. In patients with marked hyperprolactinaemia (PRL greater than 80 ng/ml or greater than 1600 mU/l), basal plasma PRL and the TRH-induced response correlated with basal plasma GH (correlation coefficients of 0.9, P less than 0.001 and 0.74, P less than 0.02, respectively). The PRL response to TRH also correlated with GH response to TRH (r = 0.38, P less than 0.01). In contrast, in patients with moderately elevated PRL (20 to 80 ng/ml), and in those with normal plasma PRL (less than 20 ng/ml or less than 400 mU/l), no such correlations were found. Immunostaining for PRL was positive in 24 out of 25 adenomas of patients with hyperprolactinaemia, while no PRL was found in the tumour tissue of 10 normoprolactinaemic patients. In conclusion, our data suggest the existence of two populations of acromegalic patients with hyperprolactinaemia, one group with correlations between GH and PRL secretion, and the other without.     

Growth hormone secretory status is a determinant of the thyrotropin response to thyrotropin-releasing hormone in euthyroid patients with hypothalamic-pituitary disease

To assess the influence of endogenous GH secretion on the TSH and T3 responses to TRH administration in patients with hypothalamic-pituitary disease, we analyzed tests in a selected group of 26 euthyroid patients with hypothalamic-pituitary disease and in 15 normal controls. Basal TSH levels and the TSH response to TRH were significantly greater in GH-deficient patients (group 1) than in patients with normal anterior pituitary function and unimpaired GH reserve (group II). However, the T3 response to TRH was significantly less in group 1 than in group II patients. In acromegaly (group III), the TSH response to TRH was blunted, while basal and stimulated T3 levels were no different compared to control levels. These findings suggest that endogenous GH depresses the TSH response to TRH while enhancing the thyroid secretion of T3 in response to the evoked TSH released.     

Differences in pituitary and testicular function between diabetic patients on insulin and oral anti-diabetic agents

Summary Pituitary responsiveness to thyrotrophin releasing hormone (TRH) and luteinizing hormone releasing hormone (LHRH) was studied in thirty one male diabetics, of whom sixteen were insulin-dependent and fifteen on oral antidiabetic agents. Ten age-matched controls were also studied. TRH and LHRH were simultaneously administered intravenously, each in a small dose of 10 g followed two hours later by 190 g and 90 g respectively. Basal hormone levels were measured in a further group of thirty six patients (twelve on insulin, twelve on oral agents and twelve on dietary restrictions alone).Higher thyrotrophin (TSH) response was observed following the small dose of TRH in the patients treated with oral agents than in the control subjects. The response of prolactin was lower in patients treated with oral agents compared with those treated with insulin. There was no difference in plasma T₃ and T₄ levels in the patients treated with insulin or oral agents. Significantly higher basal growth hormone (GH) levels were observed in the diabetics. The insulin-dependent group showed a more marked response of GH to TRH/LHRH. No response was observed in the controls. Plasma testosterone levels were significantly lower in the oral agent group (13.8 nmol/l) than in the insulin group (19.4 nmol/l), patients on dietary restrictions (18.4 nmol/l) and the control subjects (19.0 nmol/l). The LH response to the smaller dose of LHRH was impaired in patients on insulin and oral agents. There was a significant difference in FSH response between impotent and sexually normal patients.     

Paradoxical GH response to TRH during status epilepticus in man

Information on GH in relation to epilepsy is sparse, and to our knowledge there is no information on GH levels during status epilepticus in man. We studied GH in serum in six patients during status epilepticus, and in a control group of six seizure-free patients with epilepsy, before and after injection of TRH. The baseline GH values before TRH administration were within the normal range in all patients. After injection of TRH all patients with status epilepticus showed a paradoxical peak-shaped increase of GH to at least twice their baseline levels within 45 min after the injection (median basal GH value 1.5 mU/l and median peak GH value 6. 5 mU/l, mean increase 330%). No uniform reaction to TRH was observed in the control group (median basal GH value 2.7 U/l and median of the highest value within 45 min 5.2mU/l). A paradoxical peak reaction of GH to TRH was significantly more frequent in the status epilepticus group compared with the control group (P=0.008, Fisher exact probability test). TRH is not considered a GH-releasing hormone in humans during normal conditions, but a paradoxical response of GH to TRH, similar to that observed during status epilepticus, has been reported in various other pathological conditions, such as acromegaly, liver cirrhosis, mental depression and hypothyroidism. Our results of GH release after TRH administration in patients with status epilepticus suggest an altered regulation of GH as a result of the long-standing epileptic activity.     

Changes in basal and stimulated TSH and other parameters of thyroid function in acromegaly after transsphenoidal surgery

T1 and T3 levels, TSH response to TRH and somatomedin-C levels in 63 patients with acromegaly, were measured before transsphenoidal surgery and during a 4-year follow-up period. Criteria for cure were: mean GH level less than 5 mU/l, suppression of GH by oral glucose tolerance test below 2.5 mU/l and normalization of paradoxical GH reaction to TRH. Nine patients underwent radioiodine studies to assess the renal and thyroid clearance of iodide, plasma inorganic iodine level and absolute iodine uptake. Among the patients 40% had goitre, with a male preponderance. T1 and T3 levels were in the normal range both before and after surgery. A transient decrease in T3 levels was found in the immediate postoperative period. Before treatment a diminished or absent TSH response to TRH was exhibited by 64% of the goitre patients and 34% of the non-goitre groups (p less than 0.05). Despite normalization of GH and somatomedin-C levels and normal T4 and T3 levels no improvement of the TSH response was found during follow-up. No correlation between the incremental response of TSH to TRH and circulating T4 or T3 levels, basal TSH, GH or tumour size was found. There was, however, a negative correlation (r = -0.765, p less than 0.05) between the incremental TSH response to TRH and somatomedin-C levels for females with goitre. Somatomedin-C levels were higher in patients with goitre than in those without goitre (95 +/- 26 vs 75 +/- 30 nmol/l; mean +/- SD, p = 0.05). Radioiodine studies showed an increased renal clearance of iodide which was related to the increase in creatinine clearance.(ABSTRACT TRUNCATED AT 250 WORDS)     

CRITICAL STUDY OF THE GROWTH HORMONE RESPONSE TO DYNAMIC TESTS AND THE INSULIN GROWTH FACTOR ASSAY IN ACROMEGALY AFTER MICROSURGERY

Eighteen acromegalic patients were studied 8 months to 5 years after transsphenoidal selective adenomectomy in order to evaluate the most reliable criteria for assessing the efficacy of treatment. Dynamic growth hormone (GH) studies were performed after administration of oral glucose (OGTT), thyrotro-phin-releasing-hormone (TRH) or l -dopa, and after insulin-induced hypogly-caemia. Insulin-like growth factor (IGF) was measured by means of a competitive protein-binding assay which detects all somatomedins, particularly those which are GH-dependent. Of the 18 subjects reported in our study, five showed clinical symptoms of persistent acromegaly. After oral glucose, seven patients had minimum plasma GH levels above 5 ng/ml, five had GH levels between 5 and 2 ng/ml (three of whom maintained a positive GH response to TRH), and six had minimum plasma GH values below 2 ng/ml. Only the latter six patients had normal IGF levels; however, one was still responsive to TRH and three did not increase their GH plasma levels after insulin or l -dopa administration. Thus, our data suggest that the criteria used to assess the cure of acromegaly should be more rigorous than those generally applied. When plasma GH levels remain above 2 ng/ml after OGTT, it is doubtful that the cure is complete. A positive GH response to TRH after neurosurgery indicates incomplete removal of the adenoma. The lack of a GH response following insulin-induced hypoglycaemia and l -dopa administration suggests that the initial somatotroph dysfunction persists. Finally, high plasma IGF levels point to residual GH overproduction. The definition of cure in acromegaly should involve stricter biochemical criteria than those currently in use.     

Secretion of growth hormone-releasing hormone in patients with idiopathic pituitary dwarfism and acromegaly

The plasma levels of immunoreactive-GHRH in patients with idiopathic pituitary dwarfism and acromegaly were studied in the basal state and during various tests by a sensitive and specific RIA. The fasting plasma GHRH level in 22 patients with idiopathic pituitary dwarfism was 6.3 +/- 2.3 ng/l (mean +/- SD), which was significantly lower than that in normal children (9.8 +/- 2.8 ng/l, N = 21), and eight of them had undetectable concentrations (less than 4.0 ng/l). Little or no response of plasma GHRH to oral administration of L-dopa was observed in 7 of 10 pituitary dwarfs, and 3 of the 7 patients showed a response of plasma GH to iv administration of GHRH (1 microgram/kg). These findings suggest that one of the causes of idiopathic pituitary dwarfism is insufficient GHRH release from the hypothalamus. The fasting plasma GHRH level in 14 patients with acromegaly and one patient with gigantism was 8.0 +/- 3.9 ng/l, which was slightly lower than that in normal adults (10.4 +/- 4.1 ng/l, N = 72). One acromegalic patient with multiple endocrine neoplasia type I had a high level of plasma GHRH (270 ng/l) with no change in response to L-dopa and TRH test. In 3 untreated patients with acromegaly L-dopa did not induce any response of plasma GHRH in spite of inconsistent GH release, and in 4 patients with acromegaly, TRH evoked no response of plasma GHRH in spite of a marked GH release, suggesting that the GH responses are not mediated by hypothalamic GHRH.(ABSTRACT TRUNCATED AT 250 WORDS)     

Which factors predict the results of pituitary surgery in acromegaly?

Thirty-one patients with acromegaly who underwent pituitary surgery were investigated for possible predictive factors of the surgical outcome. The patients were divided into two groups: those whose GH levels normalized (less than 5 micrograms/l) after operation, group A (N = 18), and those whose GH levels remained elevated, group B (N = 13). There were no differences in age, sex distribution and sellar volume between both groups. There was a tendency to a higher incidence of supra-sellar extension of the pituitary tumor in group B (P less than 0.10). The basal GH levels in group A (38 +/- 5 micrograms/l) before operation were significantly lower than in group B (100 +/- 22 micrograms/l, P less than 0.002). Somatomedin C levels after the operation were significantly lower in group A than in group B (P less than 0.05) and were more often in the normal range (P less than 0.05). No differences were found in the occurrence of paradoxical GH responses to TRH and/or to GnRH between the groups and neither were there any differences in GH responses to GHRH and to bromocriptine between the groups. The sensitivity to SRIH tended to be higher in group A. After operation, the paradoxical GH response to TRH disappeared in 7 out of 10 patients and to GnRH in 2 out of 5 patients of group A, whereas in group B this anomaly persisted in all 9 after TRH and all 3 patients after GnRH. Hypopituitarism developed in only 5 out of the 31 patients.(ABSTRACT TRUNCATED AT 250 WORDS)     

Growth hormone response to thyrotropin-releasing hormone in diabetes

The effect of TRH on GH secretion was assessed in 13 insulin-dependent diabetics. PRL and TSH responses to TRH were also determined. Glycosylated hemoglobin levels and serial fasting glucose concentrations indicated that all but 1 of the patients had a period of poor diabetic control for several months before the study. Peak PRL and TSH levels after TRH injection in these diabetic patients did not differ significantly from values observed in nondiabetic individuals. Six of the patients responded to TRH with a significant rise in GH levels; basal GH concentrations were also elevated in these patients. Five of the 6 responders and none of the nonresponders had proliferative diabetic retinopathy. The results suggest that diabetics with elevated basal GH levels hyperrespond to TRH, and that nonspecific secretion of GH in response to TRH occurs in some patients with proliferative diabetic retinopathy. Chronic hyperglycemia does not appear to be the critical factor in determining this response.     

Altered control of growth hormone secretion in patients with cirrhosis of the liver.

Ten male patients with cirrhosis of the liver (three with portacaval anastomosis [PCA]) and eight sex- and age-matched controls underwent an arginine infusion test followed by an intravenous glucose tolerance test. Plasma glucose and growth hormone (GH) levels were measured during a period of three hours. In the normal subjects, the peak GH response to arginine occurred 60 minutes after the start of the infusion and was followed by a progressive decline in GH concentration; dextrose injection resulted in a further rapid fall in GH concentration. In cirrhotic patients, both fasting and postarginine GH concentrations were significantly higher than in controls; in addition, the dextrose injection, after causing a transitory drop in plasma GH levels, resulted in a marked increase in plasma GH concentration. In the patients with PCA, the plasma GH increase after arginine and after dextrous was more marked. In these cirrhotic patients, the plasma GH levels correlated directly with the magnitude of the portal hypertension and inversely with the serum albumin concentration, suggesting that the abnormality of GH secretion was a reflection of the derangement in liver function.     

The effect of thyrotropin-releasing hormone (TRH) and somatostatin (GHRIH) on growth hormone and prolactin secretion in vitro and in vivo in the domestic fowl (Gallus domesticus).

The effects of thyrotropin-releasing hormone (TRH) on growth hormone (GH) and prolactin (Prl) secretion have been investigated in vitro and in vivo in domestic fowl. In both conscious and anaesthetized immature chickens the administration (i.v.) of TRH (2.5 and 25 microgram/kg) significantly increased the concentration of plasma GH. The simultaneous administration of somatostatin (GHRIH), 2.5 microgram/kg, to conscious birds significantly reduced the magnitude of the GH response to TRH treatment, but had no effect on the basal levels of plasma GH. The repeated injection of TRH (10 microgram/kg) every 20 min over a 100-min period failed to maintain the concentration of plasma GH at a high level. Prl secretion was not stimulated in any of these experiments, and in anaesthetized birds TRH (2.5 and 25 microgram/kg) treatment was followed by a depression in the level of plasma Prl. The effects of TRH and GHRIH on GH secretion by an in vitro dispersed pituitary cell suspension system were very similar to the in vivo studies. TRH stimulated Prl release in vitro, in contrast to the in vivo studies, and the response was dose related. GHRIH had no effect on the basal release of Prl in vitro but significantly inhibited the response to TRH treatment.     

Paradoxical growth hormone response to thyrotropin-releasing hormone in acromegaly. Clinical correlations and prognostic value

We have investigated the effect of TRH on the release of GH in 20 acromegalic patients (14 females and 6 males) before and after selective removal of a pituitary tumour via transsphenoidal route. The follow-up period was 8 years. Pre-operatively the paradoxical response was present in 15 patients (75%). Mean GH values in TRH responders were significantly lower than in non-responders. According to the size and expansion diffusion of the adenoma, the patients were divided into 3 classes. The percentage of paradoxical response in patients in class III was significantly lower than in the other two classes. Postoperatively, mean GH values in pre-operative TRH responders were significantly lower than in non-responders; among 15 responders, 13 (86%) had postoperative GH levels under 5 micrograms/l; among 5 non-responders only 2 (40%) had GH values under 5 micrograms/l. Postoperatively 8 patients still had GH responsiveness to TRH: 6 with GH levels persistently (follow-up 8 years) under 5 micrograms/l and 2 with elevated GH values. The other 7 patients, who were responders pre-operatively and non-responders postoperatively, persistently exhibited low GH levels, except one subject who showed an increase in GH levels with reappearance of the paradoxical response, two years after surgery. These results suggest: 1. the paradoxical response may be expressed only when the hypothalamus-pituitary interactions are intact; 2. the disappearance of the paradoxical response cannot surely suggest a remission, and 3. the presence of a pre-operative paradoxical response is a good prognostic feature.     

Growth hormone response to thyrotropin releasing hormone and placebo in a group of insulin dependent diabetic patients

In order to evaluate if the assessment of the paradoxical GH responses to TRH in diabetic subjects could be altered by the presence of spontaneous fluctuations in plasma GH levels, we compared GH responses to TRH and to saline injection in 19 insulin-dependent diabetic patients. We observed significant increments (greater than 5 ng/ml) in plasma GH levels after TRH iv administration in 5 of 19 patients (26%); on the other hand, a significant increase was also observed in 6 patients (31%) after saline. We conclude that, at least in some diabetics, spontaneous GH pulses might be misinterpreted as paradoxical GH responses after a nonspecific stimulus administration. Moreover, the real existence of the paradoxical GH response to TRH has to be assessed using the following strict criteria: a sufficient magnitude in GH increment (greater than 5 ng/ml), the comparison of the kinetics of GH secretion after TRH and saline in the same patient, the presence of a significant GH rise above basal levels within the first 30 min after TRH injection.     

Evidence for autonomous secretion of prolactin in some alcoholic men with cirrhosis and gynecomastia

Prolactin responses to provocative thyrotropin releasing hormone (TRH) stimulation were evaluated in 20 cirrhotic men with gynecomastia. Fifteen of these cirrhotic men had normal responses with a minimum doubling of the prolactin concentration above basal in response to TRH. Five had abnormal (autonomous) responses in that they failed to double their basal level or had a paradoxical decrease from basal in response to TRH. Moreover, these same five men failed to have a sleep-related increase in plasma prolactin. Three of them also failed to respond to chlorpromazine stimulation. Such abnormal responses are generally associated with the presence of a prolactin secreting pituitary tumor. Basal plasma levels of prolactin were measured in all 20 men studied. The five men who failed to respond to TRH had significantly greater basal prolactin concentrations (80.5 ± 18.7 ng/ml) than did the 15 men who responded normally (33.7 ± 4.3 ng/ml) (p < 0.01), although all 20 had increased prolactin levels relative to that of controls (10.8 ± 0.9 ng/ml) (both p < 0.01).     

Growth hormone responses to oral glucose and intravenous thyrotropin-releasing hormone in acromegalic patients treated by slow-release lanreotide.

The aim of this study was to assess GH response to oral glucose tolerance test (OGTT) and TRH stimulation test in a group of 10 patients with active post-operative acromegaly before and after long-term slow-release (SR) lanreotide therapy (30 mg im every 10-14 days). Seven patients (2 males, 5 females, 29-71 yr), who during therapy maintained plasma GH and IGF-I concentrations under 5 microg/l and 450 microg/l, respectively, were considered as responders and studied for 24 (1 patient) to 36 months (6 patients). Three patients (1 male, 2 females, 46-61 yr) with levels of GH and IGF-I above those values were studied for 12 months. The OGTT (75 g po) and TRH test (400 microg iv) were repeated before and after 6, 12, 24 and 36 months. The GH response to OGTT was abnormal (nadir: >2 microg/l) at 6 and 12 months in poorly responsive patients. This response was normalized in all responsive patients. Nonetheless, 2 responsive patients showed abnormal GH values after OGTT once each throughout the 36-month study period. The GH response to TRH was characterized by great variability and exhibited unpredictable behavior throughout the study period both in responsive and in poorly responsive patients. Only 2 patients in the responsive group showed persistent normal GH levels (peak: < or =5 microg/l) after TRH for 3 yr. In conclusion, SR lanreotide treatment gave rise to a correct control of GH hypersecretion and to a normalization of GH response to oral glucose in 7 out of 10 patients, although it did not abolish the paradoxical reaction of GH to TRH in all responders. The effect of SR lanreotide on GH response to glucose tolerance test was not paralleled by GH response to TRH.     

PERSISTENT TRH-INDUCED GROWTH HORMONE RELEASE AFTER SHORT-TERM AND LONG-TERM L-THYROXINE REPLACEMENT THERAPY IN PRIMARY CONGENITAL HYPOTHYROIDISM

No appreciable changes in plasma GH levels after TRH stimulation have been observed in normal subjects, whereas acute GH release has been reported in primary hypothyroidism and other pathophysiological states. To evaluate the effect of the T4 replacement therapy on TRH-induced GH release, 28 patient volunteers with primary congenital hypothyroidism (PCH), were studied before (11 subjects), after 1 month (nine subjects) and after long-term T4 replacement therapy (eight subjects). All patients underwent a TRH test with measurement of TSH, PRL and GH levels, and were compared to 28 age-matched normal subjects. An increase of plasma GH after TRH was found in 46% of patients without any therapy, in 67% of patients after one month of T4 administration and in 75% of patients after long-term therapy. No changes were observed in plasma GH levels in controls. The TSH response to TRH was inhibited and the response of PRL was reduced step by step by T4 replacement therapy in our patients with PCH. Our results suggest that: (i) Replacement T4 therapy in PCH does not abolish the paradoxical GH response to TRH, in spite of inhibiting the TSH response and reducing the exaggerated PRL response; (ii) the GH response to TRH in PCH seems to be unrelated to low thyroid hormone levels and/or to high TSH levels, but it could be due to changes in hypothalamic-pituitary regulation which are not improved by T4 replacement therapy.