Effect of glucagon on growth hormone secretion in rats

Gentled rats injected subcutaneously with glucagon (20 microgram/100 g body weight) showed a significant decrease in plasma growth hormone (GH) at 15 min after glucagon injection. A subcutaneous injection of 50% glucose did not cause the early suppression as shown at 15 min after glucagon injection, but at 30 min after glucose injection a tendency to decrease in plasma GH was observed. In urethane anesthetized rats, a subcutaneous administration of glucagon (1 microgram or 10 microgram/100 g body weight) failed to elicit an increase in plasma GH. In vitro incubation of anterior pituitary fragments with glucagon failed to decrease the release of GH, suggesting that glucagon does not act directly on the anterior pituitary.     

Ghrelin is suppressed by glucagon and does not mediate glucagon-related growth hormone release

BACKGROUND: Glucagon stimulation is routinely used as a provocative test to assess growth hormone (GH) sufficiency in pediatrics. Ghrelin also markedly stimulates GH secretion. Because glucagon stimulates the promoter of the ghrelin gene in vitro as well as ghrelin secretion by the perfused rat stomach, we sought to determine whether ghrelin mediates glucagon-induced GH secretion. METHODS: We compared ghrelin, GH, insulin and glucose responses following administration of 0.03 mg/kg intravenously (iv; max. 1 mg) and 0.1 mg/kg intramuscularly (im; max. 2 mg) of glucagon in two groups (n = 10-11/group) of GH-sufficient children. We also measured ghrelin before and 6 min after iv administration of 1 mg glucagon in 21 adult subjects. RESULTS: In children, glucagon caused a 26% decrease in ghrelin and a 72% increase in glucose concentrations that were independent of the dose or administration route of glucagon. In contrast, the insulin response was 2-3 times higher following administration of 0.1 mg/kg im compared to 0.03 mg/kg of glucagon iv. There was a significant correlation between the maximum decrease in ghrelin and increases in glucose (p = 0.03) but not in insulin. There was a significant correlation between ghrelin and GH area under the curve after controlling for the dose of glucagon (p = 0.03) but not for the maximum increase in glucose.In normal adults, glucagon administration caused a 7% decrease in ghrelin concentrations after 6 min (p = 0.0002). CONCLUSION: Ghrelin does not play a causal role in the GH response to pharmacological glucagon administration, which suppresses ghrelin levels starting a few minutes after injection.     

Involvement of hypothalamic somatostatin in glucagon-induced suppression of growth hormone secretion in conscious rats

The role of central glucagon in regulating GH secretion was studied in conscious male rats with chronic indwelling intra-atrial and intracerebro-ventricular (ICV) cannulae. Repeated blood sampling every 20 min from 1000 hr to 1700 hr showed two major GH bursts occurring at regular intervals (3.6±0.1 hr) around 1200 hr and 1540 hr. The ICV (lateral ventricle) injection of glucagon (10 μg/rat) at 1100 hr inhibited spontaneous GH secretion, and the mean (±SE) plasma GH levels from 1120 hr to 1700 hr were lower than those in controls injected ICV with the vehicle solution only (31.9±7.8 ng/ml vs. 157.1±13.4 ng/ml, p<0.01). The GH bursts did not appear until 5 hr after the injection. The intravenous (IV) injection of glucagon (10 μg/rat) did not change plasma GH levels or the occurrence of spontaneous GH bursts. The glucagon-induced suppression of GH release was attenuated when anti-somatostatin serum (ASS), but not normal rabbit serum (NRS), was given IV in a volume of 0.25 ml immediately before the ICV injection of glucagon (10 μg/rat) (mean GH levels at 1120–1700 hr: ASS+glucagon, 133.6±26.7 ng/ml vs. NRS+glucagon, 30.5±7.4 ng/ml, p<0.01). These findings suggest that central glucagon may play an inhibitory role in regulating GH secretion by stimulating SRIF release from the hypothalamus in the rat.     

Sulfated gastrin stimulates ghrelin and growth hormone release but inhibits insulin secretion in cattle

This study was designed to determine the effects of gastrin on the circulating levels of ghrelin, growth hormone (GH), insulin, glucagon and glucose in ruminants. Two experiments were done in eight Holstein steers. Animals were randomly assigned to receive intravenous bolus injections: (1) 0.1% bovine serum albumin in saline as vehicle, 0.8, 4.0 and 20.0 μg/kg body weight (BW) of bovine sulfated gastrin-34; (2) vehicle, 0.53 μg/kg BW of bovine sulfated gastrin-17 alone or combined with 20.0 μg/kg BW of [d-Lys³]-GHRP-6, the selective antagonist of GHS-R1a. Blood samples were collected from −10 to 150 min relative to injection time. Concentrations of acyl and total ghrelin in response to gastrin-34 injection were significantly increased in a dose-dependent manner. Concentrations of GH were also markedly elevated by gastrin-34 injection; however, the effect of 20.0 μg/kg was weaker than that of 4.0 μg/kg. The three doses of gastrin-34 equally decreased insulin levels within 15 min and maintained the level until the time of last sampling. Gastrin-34 had no effect (P > 0.05) on the levels of glucagon and glucose. Levels of acyl ghrelin increased after administration of gastrin-17 alone or combined with [d-Lys³]-GHRP-6; however, [d-Lys³]-GHRP-6 did not block the elevation of GH by gastrin-17. The present results indicate that sulfated gastrin stimulates both ghrelin and GH release, but the GHS-R1a may not contribute to the release of GH by gastrin. Moreover, sulfated gastrin seems to indirectly maintain the homeostasis of blood glucose through the down-regulation of insulin in ruminants.     

Fetal and maternal endocrine responses to exercise in the pregnant ewe

Maternal and fetal concentrations of plasma insulin, pancreatic glucagon, growth hormone (GH), corticosteroids and enteroglucagon, and of blood glucose and lactate, were measured in well-fed, late pregnant ewes before, during and after walking on a treadmill at 0.7 m.s-1, 10 degrees slope for 60 min. Exercise caused rapid and substantial increases in maternal concentrations of glucose, lactate, pancreatic glucagon and corticosteroids, smaller but significant decreases in levels of GH and enteroglucagon, and no change in insulin. With the exception of GH, concentrations of these maternal hormones had returned to pre-exercise levels within 20 min of stopping exercise. The exercise-induced maternal hyperglycaemia was associated with a proportionately similar, rapid increase in fetal blood glucose; fetal blood lactate and plasma corticosteroids also increased, but at slower rates and other fetal hormone concentrations were unchanged. During recovery there was a rapid increase in fetal insulin levels. The results are discussed in terms of the regulation of exercise-induced changes in maternal energy metabolism, and fetal metabolic and hormonal sensitivity to these changes.     

Pituitary and extrapituitary effects of somatostatin in normal man.

The effects of synthetic linear somatostatin on basal circulating levels on several pituitary and pancreatic hormones, and of glucose and free fatty acids (FFA) were studied in 6 normal men after an overnight fast. A priming intravenous infusion of 250 mug of somatostatin in 18 sec was followed by a constant infusion of 500 mug over a period of 60 min. A decrease in plasma values of GH, prolactin, TSH, insulin and glucagon and in blood glucose was observed during somatostatin infusion, while FFA levels increased progressively. Plasma IRI and blood glucose increased rapidly when the somatostatin infusion was stopped, while FFA decreased progressively; GH, prolactin, TSH and glucagon remained low as compared to basal levels for one hour after the end of the infusion, i.e. until the end of the experiment. A slight but significant increase of LH and ACTH was observed after the end of the infusion.     

The Effects of Estrogen and Progesterone on the Blood Levels of Glucose,Non-Esterified Fatty Acids and Cholesterol in Ovariectomized Sheep

The effects of estrogen and progesterone on the blood levels of glucose, non-esterified fatty acids and cholesterol in ovariectomized sheep. The effects of estradiol benzoate and progesterone on blood glucose, NEFA and cholesterol were studied in ovariectomized sheep. Intramuscular injection of 2.5 mg estradiol benzoate gave biphasic changes in NEFA. After 2 hrs. NEFA was decreased, but thereafter an increase occurred and maximum levels were reached after 24 hrs. Blood glucose was significantly increased from 12 to 48 hrs. after the injection. Serum cholesterol was lowered after 24 hrs., but thereafter the level increased. Maximum values were obtained after 120 hrs. Progesterone at the same dose did not change any of the measured parameters. Simultaneous administration of estradiol benzoate and progesterone gave similar responses as estradiol benzoate alone. Blood glucose and NEFA were followed during heat in a lactating cow. Both parameters increased after ovulation. Since NEFA was increased during so long time after the injection of estradiol benzoate, the mechanism behind this effect was discussed. No lipolytic hormone has been reported to give a response of this duration. Estrogen is known to increase plasma GH, and since GH is strongly lipolytic in sheep it seemed possible that the elevated NEFA levels were caused by increased GH secretion. There is now evidence that also estrogen-induced changes in serum cholesterol are pituitary dependant. It was therefore considered possible that all the noted metabolic changes were mediated by the pituitary.     

Effect of hydrocortisone on ACTH, growth hormone, insulin and glucose in the blood of bilaterally adrenalectomized patients

This study is aimed at elucidating the mechanism of paradoxical rise in plasma ACTH levels in response to glucocorticoids, observed by several authors in bilaterally adrenalectomized patients with Cushing's disease. Six control subjects and fourteen patients bilaterally adrenalectomized for Cushing's disease were given a dose of 200 mg hydrocortisone sodium succinate by 3-5 mm i.v. injection. Plasma ACTH (in 6 patients), serum cortisol, growth hormone (GH) and insulin and blood glucose levels were estimated at 0, 30, 60, 90, and 120 minutes. The administration of hydrocortisone significantly suppressed plasma ACTH levels only at 60 min. In one case a slight rise in ACTH level during the test was observed. A significant fall in blood glucose levels was found only in the adrenalectomized patients. No significant changes in serum insulin and GH levels were noted. The possible mechanisms are discussed, especially the potential role of transient glucose deficiency in the pathophysiology of plasma ACTH increase in response to hydrocortisone in the bilaterally adrenalectomized patients.     

Effects of orally-administered melatonin and long-term exposure to light and dark on serum indices and gene expression of insulin and glucagon in male Wistar rats

We investigated influence of endogenous and exogenous melatonin on genetic and serologic aspects of secretory function of pancreas in rats. Thirty adult Wistar rats were divided into six groups. To achieve variable levels of endogenous melatonin, 10-day long-term exposure to light and darkness was implemented. Exogenous melatonin was administered orally (10 mg/kg of body weight). Blood glucose and serum levels of insulin, glucagon, and melatonin were measured by ELISA. Gene expression levels of insulin and glucagon were determined using the real time PCR. Results showed increase of blood glucose and decrease in serum levels of insulin after administration of melatonin without any significant difference in serum levels of glucagon. Gene expression levels of insulin in melatonin group were significantly lower than control group, while their glucagon was more. We concluded that oral administration of melatonin leads to increasing blood glucose, due to inhibition of insulin and stimulation of glucagon synthesis.     

The effect of hypophysectomy on pancreatic islet hormone and insulin-like growth factor I content and mRNA expression in rat

The growth arrest after hypophysectomy in rats is mainly due to growth hormone (GH) deficiency because replacement of GH or insulin-like growth factor (IGF) I, the mediator of GH action, leads to resumption of growth despite the lack of other pituitary hormones. Hypophysectomized (hypox) rats have, therefore, often been used to study metabolic consequences of GH deficiency and its effects on tissues concerned with growth. The present study was undertaken to assess the effects of hypophysectomy on the serum and pancreatic levels of the three major islet hormones insulin, glucagon, and somatostatin, as well as on IGF-I. Immunohistochemistry (IHC), in situ hybridization (ISH), radioimmunoassays (RIA), and Northern blot analysis were used to localize and quantify the hormones in the pancreas at the peptide and mRNA levels. IHC showed slightly decreased insulin levels in the cells of hypox compared with normal, age-matched rats whereas glucagon in cells and somatostatin in cells showed increase. IGF-I, which localized to cells, showed decrease. ISH detected a slightly higher expression of insulin mRNA and markedly stronger signals for glucagon and somatostatin mRNA in the islets of hypox rats. Serum glucose concentrations did not differ between the two groups although serum insulin and C-peptide were lower and serum glucagon was higher in the hypox animals. These changes were accompanied by a more than tenfold drop in serum IGF-I. The pancreatic insulin content per gram of tissue was not significantly different in hypox and normal rats. Pancreatic glucagon and somatostatin per gram of tissue were higher in the hypox animals. The pancreatic IGF-I content of hypox rats was significantly reduced. Northern blot analysis gave a 2.6-, 4.5-, and 2.2-fold increase in pancreatic insulin, glucagon, and somatostatin mRNA levels, respectively, in hypox rats, and a 2.3-fold decrease in IGF-I mRNA levels. Our results show that the fall of serum IGF-I after hypophysectomy is accompanied by a decrease in pancreatic IGF-I peptide and mRNA but by partly discordant changes in the serum concentrations of insulin and glucagon and the islet peptide and/or mRNA content of the three major islet hormones. It appears that GH deficiency resulting in a low IGF-I state affects translational efficiency of these hormones as well as their secretory responses. The maintenance of normoglycemia in the presence of reduced insulin and elevated glucagon serum levels, both of which would be expected to raise blood glucose, may result mainly from the enhanced insulin sensitivity, possibly due to GH deficiency and the subsequent decrease in IGF-I production.     

Secretion of glucagon and insulin in hypophysectomized rats: effect of tolbutamide.

Normal and hypophysectomized (hypox) rats, fed ad libitum, received intraperitoneal injections of tolbutamide (75 mg/kg/day) or of saline for 6 weeks. 24 h after the last injection, blood samples were taken for glucose, insulin and glucagon determinations. In normal rats, tolbutamide treatment did not alter serum glucose, insulin and glucagon, although it suppressed the secretion of insulin and glucagon by the pancreatic islets. In hypox rats, tolbutamide decreased serum glucose and insulin, elevated serum glucagon and stimulated the secretion of glucagon, but not that of insulin by the pancreatic islets. In addition, tolbutamide treatment increased the glucagon response to arginine in normal, but not in hypox rats. The serum glucose response to arginine was decreased by tolbutamide treatment and by hypophysectomy and, thus, appeared independent of the glucagon rise or preexisting glucagon level. We conclude that tolbutamide treatment decreased the secretion of glucagon and insulin in normal rats and stimulated that of glucagon in hypox rats, perhaps because of the low levels of insulin in the serum and in the pancreas of the latter. Our results are compatible with the hypothesis that the pancreatic action of tolbutamide is influenced by the pituitary.     

Effects of highly potent octapeptide analogs of somatostatin on growth hormone, insulin and glucagon release

Biological activities of highly potent octapeptide analogs of somatostatin (SS), D-Phe-Cys-Tyr-D-Trp-Lys-Val-Cys-Trp-NH2 (RC-160) and D-Phe-Cys-Tyr-D-Trp-Lys-Val-Cys-Thr-NH2 (RC-121), were investigated in male rats. When analog RC-160 was administered to rats in which serum growth hormone (GH) levels were elevated by pentobarbital anesthesia, a dose-related inhibition of GH was obtained at dose range of 0.1 to 2.5 micrograms/kg. The time course of GH inhibition by RC-160, RC-121 and SS-14 was studied in rats treated with phenobarbital, morphine and chlorpromazine. Analogs RC-160 and RC-121 induced a prolonged inhibition of GH levels, in contrast to SS-14, whose effect was short-lived. The analogs suppressed the GH level for more than 2 hr, the peak inhibition being seen 30 to 60 min after the injection. The effects of analogs RC-160 and RC-121 on insulin secretion were observed in rats, in which insulin levels had been elevated by intravenous administration of glucose (500 mg/rat). Administration of RC-160 suppressed insulin secretion, dose-dependently, maximum but not complete inhibition being achieved at a dose of 100 micrograms/kg. In this model, RC-160 and RC-121, in doses of 30 micrograms/kg, induced a similar inhibition of insulin release as 200 micrograms/kg of SS-14, whose action of SS-14 was transient. The effect of analog RC-160 on glucagon release was studied in rats with glucagon levels elevated by hypoglycemia. RC-160 suppressed the secretion of glucagon, the inhibition being dose-dependent in the range of 0.1 to 2 micrograms/kg. Doses of 2 and 10 micrograms/kg of this analog completely suppressed the hypoglycemia-induced glucagon release. These results indicate that analogs RC-160 and RC-121 possess prolonged and enhanced biological activities, the former analog showing a high selectivity in inhibiting GH and glucagon release in vivo as compared with that of insulin secretion.     

The release and metabolism of pancreatic hormones after major hepatectomy in the dog.

Major hepatectomy in the dog induced a 50% decrease in peripheral serum glucose, a 11-fold increase in portal plasma glucagon and a 36-fold increase in the portal glucagon/insulin ratio 3 hr after operation. Peripheral serum glucose levels were inversely correlated to the logarithmic value of portal plasma glucagon (r = -0.50, p less than 0.01) and that of the portal glucagon/insulin ratio (r = -0.85, p less than 0.01) for 1-6 hr after operation. The ratio of peripheral to portal plasma glucagon was also inversely correlated to the logarithmic value of portal plasma glucagon (r = -0.59, p less than 0.01). In case of glucose infusion, plasma glucagon levels were not elevated after major hepatectomy. The data suggest that glucose deficiency after major hepatectomy in the dog may cause hyperglucagonemia with an enhanced glucagon requirement.     

Inhibitory effects of substance P on the gamma-amino-butyric acid and gamma-hydroxybutyric acid-induced growth hormone and prolactin release in male rats

Gamma-aminobutyric acid (GABA) at 50 μg/10 μ1 was injected into the lateral ventricle after pretreatment with intraventricular injection of 1 μg of substance P in urethane anesthetized male rats. Thirty minutes after GABA injection the animals were decapitated and blood samples were collected from the trunk. Serum GH and prolactin were determined by radioimmunoassays. The intraventricular GABA elicited a significant increase in both serum GH and prolactin levels. Intraventricular substance P itself had no effect on serum GH and prolactin, but it inhibited the GABA-induced increases in serum GH and prolactin. Gamma-hydroxybutyric acid (GHB) was intraperitoneally injected with and without an intraventricular injection of substance P in urethane anesthetized rats. The GHB injection significantly increased serum GH and prolactin levels. Pretreatment with substance P completely inhibited the GHB-induced GH and prolactin responses. These results suggest that substance P might interact with GABA in the central nervous system.     

Diabetes mellitus secondary to idiopathic chronic calcifying pancreatitis in an adolescent woman

We describe the results of metabolic studies in a 17-year-old woman with diabetes mellitus which was the initial manifestation of idiopathic chronic calcifying pancreatitis (CCP). These studies were done on 2 occasions, 5 months and 5 years after the onset of diabetes, when her diabetes could be managed by glibenclamide and insulin, respectively. Five months after the onset of diabetes, oral glucose produced a small increase in insulin and a paradoxical rise in both glucagon immunoreactivity (GI) and growth hormone (GH). BY contrast, arginine-stimulated responses of the three hormones were normal. No increase in GI and a blunted rise in GH resulted from an insulin-induced decrease in blood glucose. Five years later, when CCP was demonstrated by roentogenologic examinations and tests of pancreatic exocrine function, oral glucose was followed by a flat and depressed response of C-peptide immunoreactivity and a markedly elevated response of gut glucagon-like-immunoreactivity (gut GLI). There were delayed and extremely low responses of pancreatic polypeptide to a test meal, irrespective whether or not her diabetes required treatment with insulin. These results demonstrate that CCP can cause diabetes in adolescents, as it does in adults, and that the adolescent woman described here had impaired responses of PP and gut GLI as well as insulin, GI and GH, especially to changes in blood glucose levels.     

Effect of adenosine on the serum levels of glucose, insulin and glucagon in vivo

The in vivo effect of adenosine on the serum levels of glucose, insulin and glucagon in rats fasted for twenty four hours or after an oral glucose load were studied. Under fasting conditions adenosine produced an hyperglycaemia without change in the insulin or glucagon serum levels. After a glucose load adenosine induced a marked hyperglycaemia concomitant to a decrease in insulin serum levels and an increase in glucagon serum levels. Adenosine did not alter the relationship between insulin and glucagon. In vivo adenosine administration altered the secretion of hormones by the islets of Langerhans (increased the release of glucagon and decreased the secretion of insulin) but this was only clearly observable under stimulated conditions. Adenosine did not alter the regulatory mechanism(s) that modulate the relationship between insulin and glucagon.     

Growth hormone and prolactin secretion after metoclopramide administration (DA2 receptor blockade) in fertile women.

In this report, we will describe the results of a cross-sectional study to assess PRL and GH secretion during the early follicular phase in 22 fertile patients after metoclopramide administration in order to achieve a dopaminergic DA2 receptor blockade. Blood samples were collected at - 15, 0, 15, 30, 45 and 60 minutes. PRL, GH, estradiol, IGF-I, TSH, glucose, and insulin were measured in the samples taken at - 15 and 0 minutes. The existence of a correlation between GH and PRL secretion was investigated. All patients presented normal serum levels of estradiol, prolactin, insulin, fasting glucose and IGF-I. Serum GH levels were not changed after metoclopramide infusion (p = 0.302), but there was a significant alteration in serum PRL (p = 0.0001) with the highest levels after 30 (mean: 237.20 ng/ml +/- 95.86) and 45 (mean: 211.80 ng/ml +/- 83.24) minutes. Serum GH levels did not correlate with serum PRL levels after the dopaminergic DA2 blockade. We conclude that GH secretion was not modulated by a direct effect of type 2 dopamine receptor.     

D-Trp8-D-Cys14-somatostatin--demonstration of its differential suppressive activity in juvenile diabetics.

In 8 insulin-dependent diabetics, the effect of D-Trp8-D-Cys14-somatostatin on blood glucose, growth hormone, and glucagon levels as well as on insulin requirements from an artificial endocrine pancreas was studied during a balanced meal. The somatostatin analogue was infused at a rate of 25 microgram/h preceeded by a bolus injection of 25 microgram 30 minutes before ingestion of the meal. At this dose the analogue had no effect on glucagon levels and insulin requirements from the artificial pancreas. On the other hand, there was a significant lowering effect on fasting blood glucose levels, possibly indicating a direct inhibition of hepatic glucose production. Furthermore, there might be a slight effect on growth hormone levels, as was demonstrated by a rebound increase after termination of analogue infusion.     

Ghrelin directly stimulates glucagon secretion from pancreatic alpha-cells

Previous work has demonstrated that the peptide hormone ghrelin raises blood glucose. Such has been attributed to ghrelin's ability to enhance GH secretion, restrict insulin release, and/or reduce insulin sensitivity. Ghrelin's reported effects on glucagon have been inconsistent. Here, both animal- and cell-based systems were used to determine the role of glucagon in mediating ghrelin's effects on blood glucose. The tissue and cell distribution of ghrelin receptors (GHSR) was evaluated by quantitative PCR and histochemistry. Plasma glucagon levels were determined following acute acyl-ghrelin injections and in pharmacological and/or transgenic mouse models of ghrelin overexpression and GHSR deletion. Isolated mouse islets and the α-cell lines αTC1 and InR1G9 were used to evaluate ghrelin's effects on glucagon secretion and the role of calcium and ERK in this activity. GHSR mRNA was abundantly expressed in mouse islets and colocalized with glucagon in α-cells. Elevation of acyl-ghrelin acutely (after sc administration, such that physiologically relevant plasma ghrelin levels were achieved) and chronically (by slow-releasing osmotic pumps and as observed in transgenic mice harboring ghrelinomas) led to higher plasma glucagon and increased blood glucose. Conversely, genetic GHSR deletion was associated with lower plasma glucagon and reduced fasting blood glucose. Acyl-ghrelin increased glucagon secretion in a dose-dependent manner from mouse islets and α-cell lines, in a manner requiring elevation of intracellular calcium and phosphorylation of ERK. Our study shows that ghrelin's regulation of blood glucose involves direct stimulation of glucagon secretion from α-cells and introduces the ghrelin-glucagon axis as an important mechanism controlling glycemia under fasting conditions.     

Effect of thyrotropin releasing hormone injection on blood growth hormone (GH), TSH and growth hormone releasing hormone (GHRH) concentrations in cancer patients

In order to investigate whether endogenous GHRH and somatostatin were involved in the mechanism of the paradoxical GH rise after TRH injection, changes in serum GH and plasma GHRH were examined before and after TRH injection in 12 cancer patients and changes in serum TSH and GH were similarly studied in 76 cancer patients including 31 GH-responders and 45 GH-nonresponders to TRH. TRH stimulated GH secretions without altering the circulating GHRH concentration in 4 of the 12 cancer patients. There was neither a significant correlation between the increase from the basal to maximum GH and GHRH after TRH injection in the 12 cancer patients nor a reciprocal relationship between the increase in GH and TSH after TRH injection in the 76 cancer patients. These findings suggested that the paradoxical GH rise after TRH injection in cancer patients was exerted by its direct action at the pituitary level, and not mediated through the hypothalamus.