Growth Hormone-producing Pituitary Adenomas Research Articles

Co-localization of human growth hormone and human prolactin in hormonally pure human growth hormone-producing pituitary adenomas.

Co-localization of human growth hormone (hGH) and human prolactin (hPRL) in hGH-producing pituitary adenomas was examined by electron microscopy with immunoblot analysis. At the electron microscopic level using anti-hGH or anti-hPRL polyclonal antibody, hGH and hPRL were found to be co-localized within each of the secretory granules in one of five cases. Double-labeling electron immunocytochemistry using colloidal gold particles of different sizes was effective in demonstrating this co-localization. As an additional step, we performed immunoblot analysis of hGH-producing pituitary adenomas using monoclonal antibodies. Four hGH-producing adenomatous tissue samples contained several hPRL-immunoreactive bands. In Case 2, the main 23K hPRL band was stained especially strongly The immunoblotting analysis of purified hGH using both anti-hPRL polyclonal antibody and monoclonal antibody to asses cross-reaction of the polyclonal anti-hPRL antisera with hGH revealed that both monoclonal and polyclonal antibodies were suitable for determining the co-localization. Double-labeling techniques using anti-hGH and anti-hPRL monoclonal antibodies demonstrated that only a few secretory granules were positive for co-localization of both hGH and hPRL (Case 2). The present study, which used not only polyclonal but also monoclonal antibodies, suggests that some hGH-producing pituitary adenomas contained both hPRL and hGH in the same secretory granules of tumor cells.

Radiation Therapy Alone for Growth Hormone-Producing Pituitary Adenomas

We present our experience in the treatment of growth hormone (GH)-producing pituitary adenomas using irradiation alone. Between 1983 and 1991, 21 patients suffering from GH-secreting pituitary adenomas were treated with radiotherapy alone. Two bilateral opposing coaxial fields were used in 10 patients and in the remaining 11 a third frontovertex field was added. Treatment was given in 1.8-2 Gy daily fractions and total dose ranged between 45 and 54 Gy. Treatment was given using a cobalt unit. Four patients treated with somatostatin prior to and 14 patients treated after the end of radiotherapy experienced symptom relief for 6-28 weeks. The 5-year actuarial rate of disease control was 72%. Five out of six failed patients had macroadenomas. Hypopituitarism was observed in 5/21 (24%) patients. Whereas RT alone is effective in the treatment of microadenomas, this is not true for large infiltrative macroadenomas.

Somatostatin receptor subtypes sst1, sst2, sst3 and sst5expression in human pituitary, gastroentero‐pancreatic and mammary tumors

Using in situhybridization techniques with selective oligoprobes, the gene expression of sst1, sst2, sst3 and sst5 was studied in a series of 32 human pituitary adenomas, 28 breast tumors and 21 endocrine gastroentero-pancreatic tumors, shown to express somatostatin receptors to variable extents. In most of these tumors the sst2 receptor subtype was abundantly expressed, even though a significant number of pituitary adenomas, breast and gastroentero-pancreatic tumors expressed sst1 and/or sst3 as well. A very high incidence of the sst5subtype was found in growth hormone-producing pituitary adenomas and, to a lesser extent, in inactive pituitary adenomas, whereas breast tumors seldom expressed sst5; gastroentero-pancreatic tumors showed all possible combinations of sst expression, with, however, a predominance of sst2 and sst1. Overall, the presence of sst2 mRNA and/or sst5 mRNA generally correlated with the presence of octreotide binding sites. A lack of octreotide binding sites corresponded with a lack of sst2 mRNA. Several tumors exhibiting a low number of octreotide binding sites had no measurable sst2 mRNA, despite abundance of β-actin mRNA, suggesting in these cases a very low abundance of sst mRNAs or a too low sensitivity of the in situ hybridization methodology. In all other cases, the method allowed precise localization of the respective mRNAs on the tumor tissue, notably in breast tumors with non-homogeneous receptor distribution. Tumors without measurable amounts of somatostatin receptors had no detectable sst mRNA. Our results indicate a highly variable abundance of the various sst mRNAs in individual somatostatin receptor-containing tumors. Int. J. Cancer 70:530–537. © 1997 Wiley-Liss Inc.

Somatostatin receptor subtypes sst1, sst2, sst3 and sst5 expression in human pituitary, gastroentero-pancreatic and mammary tumors: comparison of mRNA analysis with receptor autoradiography.

Using in situ hybridization techniques with selective oligoprobes, the gene expression of sst1, sst2, sst3 and sst5 was studied in a series of 32 human pituitary adenomas, 28 breast tumors and 21 endocrine gastroentero-pancreatic tumors, shown to express somatostatin receptors to variable extents. In most of these tumors the sst2 receptor subtype was abundantly expressed, even though a significant number of pituitary adenomas, breast and gastroentero-pancreatic tumors expressed sst1 and/or sst3 as well. A very high incidence of the sst5 subtype was found in growth hormone-producing pituitary adenomas and, to a lesser extent, in inactive pituitary adenomas, whereas breast tumors seldom expressed sst5; gastroentero-pancreatic tumors showed all possible combinations of sst expression, with, however, a predominance of sst2 and sst1. Overall, the presence of sst2 mRNA and/or sst5 mRNA generally correlated with the presence of octreotide binding sites. A lack of octreotide binding sites corresponded with a lack of sst2 mRNA. Several tumors exhibiting a low number of octreotide binding sites had no measurable sst2 mRNA, despite abundance of beta-actin mRNA, suggesting in these cases a very low abundance of sst mRNAs or a too low sensitivity of the in situ hybridization methodology. In all other cases, the method allowed precise localization of the respective mRNAs on the tumor tissue, notably in breast tumors with non-homogeneous receptor distribution. Tumors without measurable amounts of somatostatin receptors had no detectable sst mRNA. Our results indicate a highly variable abundance of the various sst mRNAs in individual somatostatin receptor-containing tumors.

Morphological effects of octreotide on growth hormone-producing pituitary adenomas

Morphological effects of octreotide on growth hormone-producing pituitary adenomas

Cytokeratin distribution and functional properties of growth hormone-producing pituitary adenomas.

In addition to its structural function, cytokeratin may have other important roles within cells. We have reported that in growth hormone-producing adenomas (GH cell adenomas), two distinct types can be recognized by their cytokeratin distribution patterns (dot-like or perinuclear pattern) and that each type has different clinicopathological and endocrinological properties. To confirm these phenomena in a larger series and to clarify the significance of different cytokeratin distribution patterns, we studied cytokeratin localization in 70 GH cell adenomas from acromegalic patients. Type I adenomas ( 15) almost exclusively (>98%) composed of cells with a prominent, dot-like distribution; type 2 adenomas (36) comprised of cells with perinuclear cytokeratin; and type 3 adenomas (11) comprised of both cell types were separated. The remaining 8 did not exhibit a distinct distribution pattern. By electron microscopic immunocytochemistry for cytokeratin, dot-like distribution corresponded to fibrous bodies, whereas perinuclear distribution represented immune deposition in the perinuclear zone. Immunohistochemistry for GH, prolactin, β-thyrotropin, and α-subunit of glycoprotein hormones revealed a reduced expression of these hormones in type 1 adenomas, compared with types 2 and 3 adenomas. In normal pituitary glands, almost all GH cells showed a perinuclear cytokeratin distribution, and only a few GH cells exhibited a dot-like pattern. These findings suggest that a dot-like cytokeratin distribution in GH cells may be pathological (a change from physiological perinuclear distribution) and that adenomas with such a distribution may reduce endocrine activities as a result of unknown factors.

Growth hormone-producing pituitary adenomas: MR characteristics and pre- and postoperative evaluation.

Twelve patients with growth hormone-producing pituitary adenomas were investigated to evaluate correlations between magnetic resonance (MR) appearance and tissue characteristics, sellar expansion, suprasellar tumor extension, and tumor consistency. MR imaging could detect microadenomas with a characteristic enhancement delay of gadolinium-diethylenetriaminepenta-acetic acid in the surrounding normal pituitary gland. Macroadenomas and giant tumors appeared on T2-weighted images as mostly high intensity, while varied greatly on T1-weighted images. The signal intensities of growth hormone-producing adenomas were more homogeneous (80%) than those of other functioning and non-functioning adenomas on T1-weighted images. However, there was no typical MR imaging for the tissue characteristics of the growth hormone-producing adenomas, and no particular correlation with suprasellar extension or consistency. Postoperative MR findings showed that the transcranial approach was superior for total or subtotal removal of large tumors, especially with suprasellar extension up to the upper third ventricle. The transsphenoidal approach can achieve a satisfactory decompressive effect on the optic chiasm, but more radical removal is required to achieve a normal hormone level.

Expression of Glycoprotein Hormones and Intracytoplasmic Distribution of Cytokeratin in Growth Hormone-Producing Pituitary Adenomas

Sixteen growth hormone (GH)-producing pituitary adenomas were studied for the expression of glycoprotein hormone subunits and cytokeratin by light microscopic immunohistochemistry. Cytokeratin immunoreactivity was demonstrated in all adenomas, but its intracytoplasmic distribution showed two distinct patterns; a prominent, dot-like pattern and a diffuse, perinuclear pattern. Seven adenomas (type 1) were exclusively composed of cells with cytokeratin in a dot-like pattern, whereas 9 adenomas (type 2) comprised of cells with cytokeratin of perinuclear distribution. The expression of alpha-subunit of glycoprotein hormone was significantly different between the two types of adenomas; 8 of 9 adenomas of type 2 contained many alpha-subunit immunoreactive cells but none of type 1 adenomas showed any immunoreactivity. Only a small number of adenoma cells were positive for beta-subunit of thyrotropin stimulating hormone in 3 adenomas of type 2. beta-subunits of follicle stimulation hormone and luteinizing hormone were negative in all adenomas. These findings suggest that the expression of glycoprotein hormone subunits in GH-producing adenomas may be closely linked to their types distinguishable by the cytokeratin distribution pattern.

Bromocriptine-induced Morphological Changes in Cultured Growth Hormone producing Pituitary Adenoma Cells

Tumor tissues obtained from two patients with growth hormone-producing pituitary adenomas were cultured and treated with bromocriptine, then examined for morphological changes. Untreated tumor cells (controls) were morphologically well preserved and in terms of growth hormone secretion. Tumor cells exposed to bromocriptine for 7 days contained many vacuoles, which, after 14 days of exposure, grew larger and more numerous. These vacuoles appeared to be extensions of endoplasmic reticulum: some were connected to rough or smooth endoplasmic reticulum or, occasionally, to Golgi apparatus, and there were ribosomes on their surfaces. Bromocriptine apparently has cytocidal effects on growth hormone-producing pituitary adenoma in vitro.

Hypothalamic Neuronal Hamartoma and Adenohypophyseal Neuronal Choristoma

Hypothalamic neuronal hamartomas and neuronal choristomas of the anterior pituitary are rare lesions; either may be associated with endocrinopathy. We describe a case of each with associated growth hormone-producing pituitary adenomas and clinical acromegaly, both well documented and studied by immunocytochemistry and electron microscopy. That a functional relationship exists between the neuronal malformation and the pituitary neoplasm remains speculative. We suggest that a growth hormone-releasing factor-like substance may have been elaborated by hypothalamic-type neurons, which, by a trophic effect, may have resulted in the production of an adenohypophyseal neoplasm. Our study supports the concept that secretory neurons, either outside or within the sella, may induce adenomas.

Prolactin and growth hormone-producing pituitary adenomas. An immunohistochemical and ultrastructural study.

Thirteen prolactin and five growth hormone-producing pituitary adenomas were studied by immunohistochemistry and electron microscopy. The immunohistochemical localization of prolactin and growth hormone correlated well with elevated serum levels of pituitary hormones in all cases. Ultrastructural characterization by granule density and secretory activity was studied in relation to the serum levels of pituitary hormones and the sizes of the tumors. This indicated that markedly elevated serum hormone levels were related to larger tumors with high secretory activity, as indicated by abundant endoplasmic reticulum and well-developed Golgi complexes rather than to the numbers of cytoplasmic granules in the tumors. Two patients with prolactin-producing adenomas had been treated with bromocryptine before surgery. Both tumors showed evidence of degeneration, including cytoplasmic vacuolization. In one case the tumor had an increased number of secretory granules, while in the other case there were few viable cells and an abundance of amyloid deposits. The effects of bromocryptine therapy on pituitary tumor morphology in these two cases include an increased number of pituitary granules, and cellular degenerative changes.

Use of immunohistochemical and morphologic methods for the identification of human growth hormone-producing pituitary adenomas.

Indirect immunoperoxidase and immunofluorescent techniques were used for the detection of growth hormone in routinely processed human pituitary adenomas. With the immunoperoxidase method, distinct immunoreactive cells could be demonstrated in 15 adenomas; all except for one were associated with acromegaly. Immunofluorescence was less useful because of autofluorescence due to the fixation. In histologic stainings, the immunoreactive cells were acidophyl. With immunohistochemistry and electron microscopy, densely granulated and sparsely granulated adenomas could be distinguished. Sparse granulation was associated with an expanded rough-surfaced endoplasmic reticulum and Golgi complex. Clinically, the sparsely granulated adenomas were characterized both by active hormone secretion and aggressive local growth. It is concluded that the immunoperosidase method is of value for functional classification of pituitary adenomas. For information about secretory activity we suggest that electron micrographs of adenoma cells be examined.