Histologic and immunohistochemical study of clinically non-functioning pituitary adenomas

Seventy-five clinically non-functioning pituitary adenomas were characterized in terms of their histologic and immunohistochemical features. Fourteen adenomas (18.7%) were positive for hormones other than gonadotropins. These included two somatotroph adenomas, three lactotroph adenomas, four thyrotroph adenomas, four corticotroph adenomas and one unusual plurihormonal adenoma. Fifty-five adenomas (73.3%) were exclusively positive for one or more gonadotropin subunits (β-follicle stimulating hormone, p-luteinizing hormone, and α-subunit of glycoprotein hormones), but negative for other hormones such as growth hormone and β-thyrotropin. Histologically, a papillary arrangement along the capillary was most characteristically observed in the gonadotropin-positive adenomas. Five of six adenomas negative for any pituitary hormones exhibited the typical papillary structure. Thus, approximately 80% of clinically non-functioning adenomas constituted a single tumor group that shared the common histologic features of gonadotroph adenomas. These findings suggest that nearly all tumor types of clinically non-functioning adenomas can be diagnosed solely by light microscopy in combination with immunohistochemistry, and that the vast majority of them may be gonadotroph adenomas.     

Plurihormonal pituitary adenomas

Plurihormonal adenomas of the pituitary, ie, tumors that engage in the production of unusual combinations of hormones, represent approximately 10% to 15% of all adenomas. Such tumors comprise in excess of 50% of adenomas in the setting of acromegaly and occur with somewhat greater frequency in childhood and adolescence than in adulthood. Eight percent are associated with multiple endocrine neoplasia, type I. The most common variant of plurihormonal adenoma produces growth hormone, prolactin, and one or more glycoprotein hormones, the most common being TSH. Clinical effects most often reflect the presence of growth hormone, and to a lesser extent, prolactin cells; expression of glycoprotein hormone production is rare. The tumors are more often macroadenomas (80%) than microadenomas (20%) and demonstrate gross invasion in 50% of cases. Plurihormonal adenomas may be ultrastructurally monomorphous, bimorphous, or trimorphous; thus, one morphologic cell type may elaborate several hormones.     

CDKN2A/p16 inactivation is related to pituitary adenoma type and size

p16 (CDKN2A, MTS1, INK4A) status at genomic and protein levels was analysed and correlated with clinico-pathological features in 72 pituitary adenomas. Methylation of CpG islands of promoter/exon 1 sequences was found in most gonadotroph, lactotroph, plurihormonal, and null cell adenomas (36 of 44, 82%), but it was rare in somatotroph (1 of 13 cases, 8%) and corticotroph adenomas (1 of 15 cases, 7%). Homozygous CDKN2A deletion was restricted to rare somatotroph (15%) and corticotroph adenomas (13%). Immunohistochemical p16 protein expression was observed in the normal adenohypophysis, whereas it was absent in 60 of 72 (83%) tumours and reduced in another ten (14%) tumours. Staining for p16 was only seen in 5 of 15 (33%) corticotroph, 3 of 13 (23%) somatotroph, 3 of 5 (60%) plurihormonal, and 1 of 19 (5%) null cell adenomas. p16 immunonegativity without CDKN2A methylation or deletion occurred in 22 tumours, including most somatotroph and corticotroph adenomas (15 of 28, 54%). Both CDKN2A alterations and p16 negativity were related to larger tumour size. Patients with p16-negative tumours were older than patients with p16-positive tumours. These data suggest that p16 down-regulation is common in all adenoma types. The mechanisms of p16 down-regulation probably involve CDKN2A methylation in most types, but remain to be determined in somatotroph and corticotroph adenomas. These findings also suggest that p16 down-regulation is usually not an initial event, but is acquired during adenoma progression.     

Prolactin-containing pituitary adenomas. Their characteristics and comparative study with non-prolactin adenomas.

Immunohistochemical study of 130 pituitary adenomas shows that 31% are prolactin-containing adenomas, two-thirds of which are monohormonal adenomas, i.e. prolactin cell adenoma, and one-third are multihormonal adenomas, i.e. mixed growth hormone cell-prolactin cell adenoma and plurihormonal adenoma with prolactin. Clinical symptoms including amenorrhea and galactorrhea are not useful in distinguishing prolactin from non-prolactin adenomas. Serum prolactin concentration of 80 ng/ml is a good cut-off point to distinguish prolactin cell adenoma from non-prolactin adenoma but can not separate many of the multihormonal adenomas from non-prolactin adenomas. Calcification is not only more commonly seen but also more prominent in prolactin-containing adenomas. Spheroid amyloid is present in one prolactin cell adenoma. Immunohistochemistry is specific and reliable in identifying prolactin-containing adenomas. All prolactin cell adenomas and 2/13 multihormonal adenomas show paranuclear staining of prolactin in almost every adenoma cell. The remaining (11/13) multihormonal adenomas show less prolactin cells and diffuse cytoplasmic staining of prolactin. The prolactin staining pattern in the latter group is unique and appears to be indicative of the presence of other hormone(s).     

Immunohistochemical study of p53 protein in human and animal pituitary tumors

In many human cancers, p53 gene mutations are frequently occurring genetic abnormalities, which may be detected by immunohistochemical staining for p53 protein. In the present study, p53 immunoreactivity was investigated in formalin-fixed, paraffin-embedded tissues from human and animal pituitary tumors, using the avidin-biotin-peroxidase complex technique. No p53 was detected in 3 nontumorous human adenohypophyses or in 40 human pituitary tumors including 5 GH cell adenomas, 10 PRL cell adenomas, 2 mixed GH cell-PRL cell adenomas, 2 acidophil stem cell adenomas, 8 ACTH cell adenomas, 1 TSH cell adenoma, 1 FSH/LH cell adenoma, 5 null cell adenomas, 5 oncocytomas, and 1 plurihormonal adenoma. Twenty nontumorous and hyperplastic pituitaries of hGRH transgenic mice and 8 tumors in these transgenic animals were immunonegative for p53. All pituitary tumors found in AVP/SV40 transgenic mice contained p53 immunoreactivity in the nuclei, while the nontumorous adenohypophysis of one such transgenic mouse was negative. It can be concluded that p53 mutations are apparently not involved in the pathogenesis of human pituitary adenomas or of the pituitary tumors which develop in hGRH transgenic mice. However, pituitary tumors in AVP/ SV40 transgenic mice are accompanied by p53 expression.     

Double pituitary adenomas: a series of three patients

Multiple pituitary adenomas may occur in up to 1.6-3.3% of patients with Cushing's syndrome. We report three patients with double pituitary adenomas detected at surgery. Two patients had Cushing's disease, but trans-sphenoidal exploration revealed a small prolactinoma in each. One prolactinoma also contained small numbers of basophils. Re-operation in both patients because of persistent Cushing's syndrome showed an ACTH-secreting micro-adenoma. The third patient with acromegaly had two macro-adenomas discovered in different parts of the gland at surgery: one plurihormonal and one null cell tumour. Careful evaluation of pre-operative MRI may not always detect more than one pituitary adenoma.     

Localization of insulin-like growth factor-II mRNA in human pituitary adenomas

Insulin-like growth factors (IGFs) have been reported to promote cell proliferation in many tumours, but their contribution to pituitary adenoma development and growth has not been characterized. We report the presence of insulin-like growth factor II (IGF-II) mRNA in pituitary adenomas using in situ hybridization (ISH). The intensity of IGF-II hybridization signal was correlated with adenoma type, and the presence of Ki-67. Among the 109 adenomas examined, 55 (50.4%) were positive for IGF-II mRNA. All acidophil stem cell, functioning corticotrophic and plurihormonal adenomas contained the message; a high incidence of signal was found among sparsely (7/8) and densely (4/6) granulated growth hormone (GH) cell adenomas, mixed GH cell–prolactin (PRL) cell adenomas (6/7), thyrotrophic (4/6) and null-cell (6/7) adenomas. Less frequently, IGF-II mRNA was localized in mammosomatotrophic, silent subtype 3, gonadotrophic, and oncocytic adenomas, whereas all sparsely granulated PRL cell adenomas and silent corticotrophic adenomas of subtypes 1 and 2 were negative. The MIB-1 labelling index was significantly higher in adenomas with a moderate to intense IGF-II signal than in adenomas with weak or no signal. The results suggest that IGF-II, when highly expressed, may have a role in pituitary adenoma proliferation.     

Contribution of immunohistochemistry, electron microscopy, and cell culture to the characterization of nonfunctioning pituitary adenomas: A study of 40 cases

We studied 40 endocrinologically inactive pituitary adenomas by immunohistochemistry, electron microscopy, and cell culture in order to determine the incidence of gonadotropic adenomas and to classify nonfunctioning adenomas. Immunohistochemical studies using a large panel of monoclonal and polyclonal antibodies identified the following nonfunctioning adenomas: 20 gonadotropic adenomas, four silent corticotropic adenomas, one plurihormonal adenoma, and 15 nonsecreting adenomas. Among nonsecreting adenomas, ultrastructural study of 13 cases identified seven null cell adenomas and six oncocytomas. Silent corticotropic adenomas were classified into subtypes I, II, and III according to Kovacs and Horvath. Most often, gonadotropic adenomas displayed a varying number of oncocytic cells, characteristic secretory granules, and a prominent Golgi apparatus. Postembedding immunoelectron microscopy was performed on eight gonadotropic or nonsecreting adenomas, but this technique did not provide any additional information. Six gonadotropic adenomas and 10 so-called nonsecreting adenomas were studied in primary cell cultures. The six gonadotropic adenomas and seven of the 10 nonsecreting adenomas released gonadotropins in the culture medium. The use of in vitro results as a supplementary diagnostic criterion allowed classification of the 40 nonfunctioning adenomas as follows: 27 gonadotropic adenomas, four silent corticotropic adenomas, one plurihormonal adenoma, and eight nonsecreting adenomas. These results demonstrate a high proportion of gonadotropic adenomas among nonfunctioning adenomas (67.5%) and the usefulness of several techniques in characterizing this type of pituitary adenoma.     

Localization of prolactin in chromophobe pituitary adenomas: study of human necropsy material by immunoperoxidase technique.

In order to identify prolactin-producing tumours in human pituitary glands, 45 chromophobe adenomas, obtained from unselected necropsies, have been studied by various staining procedures including the immunoperoxidase technique for the demonstration of prolactin. The presence of immunoreactive prolactin was revealed in the cytoplasm of the tumour cells in six cases (13%), indicating that the occurrence of prolactin-producing adenomas is not rare. No correlations were established between tumours and clinical history. Two adenomas were detected in female and four in male patients. The age of the patients at necropsy ranged from 28 to 75 years. Three adenomas were associated with disseminated carcinoma, two with fatal liver disease, and one with diabetes mellitus, atherosclerosis, and pyelonephritis. Manifest endocrine symptoms were not disclosed, and endocrine investigations, including measurements of blood prolactin levels, were not undertaken. Thus, direct evidence is lacking as to whether or not these tumours were actively secreting prolactin. In the non-tumorous parts of the anterior lobes the number of prolactin cells was decreased in two cases, suggesting that prolactin released from the adenoma cells suppressed prolactin production in the non-tumorous pituitary. However, the number of prolactin cells of the non-tumorous adenohypophysis seemed to be unchanged in two and increased in another two cases. The present findings conclusively proved the existence of the prolactin-producing adenomas as a distinct entity. These tumours do not stain with acid or basic dyes, they are PAS or thionin negative, and do not contain immunoreactive growth hormone. Thus, by conventional staining procedures they are indistinguishable from other chromophobe adenoma types. Herlant's erythrosin and Brookes' carmoisine methods, claimed spedifically to stain prolactin cells, failed to provide reliable results, hence their use cannot be recommended in tumour identification. Immunoperoxidase staining of prolactin is the only technique which conclusively reveals the presence of immunoreactive prolactin in the cytoplasm of the tumour cells and permits diagnosis. It is proposed that this technique be introduced in pituitary morphological studies. Its application may lead to a better understanding of problems related to prolactin-producing tumours and their secretory activity.     

Expression of interleukin-6, interleukin-6 receptor (gp80), and the receptor's signal-transducing subunit (gp130) in human normal pituitary glands and pituitary adenomas.

Interleukin-6 (IL-6) is an important cytokine in cell proliferation and differentiation in several organs. It has also been reported that IL-6 plays a role in secretion or release of pituitary hormones in pituitary hormone-secreting cells and pituitary adenomas, but convincing data in situ have not yet been reported. In this study, we examined the participation of IL-6 in the production of pituitary hormones and the differences between human normal pituitary glands and pituitary adenomas by determination of the localization or expression of IL-6, IL-6 receptor (IL-6R, gp80), and the signal-transducing subunit (gp130) of the receptor using immunohistochemical staining and RT-PCR. IL-6 was mainly expressed in ACTH- and FSH/LH-secreting cells in normal pituitary glands, as shown by double staining. gp 80 and gp130 were coexpressed in almost all GH- and PRL-secreting cells and in approximately 30% of FSH/LH-secreting cells. RT-PCR showed that IL-6 mRNA was expressed in only one of all the pituitary adenomas examined, whereas gp 80 and gp 130 mRNAs were detected in all these pituitary adenomas. In conclusion, IL-6 was mainly expressed in ACTH- and FSH/LH-secreting cells, and the receptors were expressed in GH-, PRL- and FSH/LH-secreting cells in human normal pituitary glands. Furthermore, our data emphasized that the mechanism of IL-6 function in human pituitary adenoma cells is distinct from that in normal pituitary cells.     

Alpha-subunit immunoreactivity in plurihormonal pituitary adenomas of patients with acromegaly.

Twelve plurihormonal pituitary adenomas, removed surgically from 11 of 20 patients with acromegaly, were investigated. The mean age of the 11 patients was 45 yr. Seven patients with eight tumors were men. The tumors were immunostained for all known adenohypophysial hormones by the avidin-biotin-peroxidase complex (ABC) technique. All 12 adenomas were immunoreactive for growth hormone (GH) and alpha-subunit; prolactin (PRL) was detected in five tumors. Stains for the beta-subunits of glycoprotein hormones [thyroid-stimulating hormone (TSH), follicle-stimulating hormone (FSH), and luteinizing hormone (LH)] revealed immunopositivity for specific beta-subunits in ten adenomas. The close association between GH and alpha-subunit in pituitary tumors of acromegalic patients remains to be explained.     

Immunohistological studies on clinically silent pituitary adenomas

Clinically inactive adenomas represent 30.7% of all pituitary tumors in our surgical collection of 616 cases. Ninety-six tumors were studied immunohistologically with many antibodies for their hormone content. Morphological classification of these adenomas reveals oncocytic adenomas in 42 cases (44%), small cell chromophobe adenomas in 33 (34%), large cell chromophobe adenomas in 14 cases (15%), undifferentiated mucoid cell adenomas in 4 cases (4%) and undifferentiated acidophil adenomas in 3 cases (3%). Immunohistological studies performed with the six pituitary hormones GH, prolactin, ACTH, TSH, LH and FSH and additionally with α-subunit demonstrated nearly all possible combinations of hormones in adenoma cells. The most frequently occurring (29%) was LH (in 3% of adenomas alone); α-subunit followed in frequency (24%), with FSH present in 21%. Combinations of 2 hormones were detected in 16%, of 3 in 13% and of more than 3 hormones in 2%. All 6 hormones and α-subunit were negative in 48% of adenomas. It must be concluded 1) that many clinically silent adenomas are LH- or FSH- or α:-subunit-positive and therefore probably originate from gonadotropic cells, and 2) that clinically silent adenomas of acidophil cell type or mucoid cell type are rare. Although many of these adenomas apparently do not secrete the hormone which they immunohistologically contain, determining the plasma levels of the gonadotropins, and especially the α-subunit in clinical studies, may obtain a suitable and helpful clinical marker in the diagnosis of “endocrine inactive” adenomas, and especially of their recurrences.     

My approach to pathology of the pituitary gland

The sellar region is the site of a large number of pathological entities arising from the pituitary and adjacent anatomical structures, including brain, blood vessels, nerves and meninges. The surgical pathology of this area requires the accurate identification of neoplastic lesions, including pituitary adenoma and carcinoma, craniopharyngioma, neurological neoplasms, germ cell tumours, haematological malignancies and metastases, as well as non-neoplastic lesions such as cysts, hyperplasias and inflammatory disorders. This review provides a practical approach to the diagnosis of pituitary specimens that are sent to the pathologist at the time of surgery. The initial examination requires routine haematoxylin and eosin staining to establish whether the lesion is a primary adenohypophysial proliferation or one of the many other pathologies that occurs in this area. The most common lesions resected surgically are pituitary adenomas. These are evaluated with several special stains and immunohistochemical markers that are now available to accurately classify these pathologies. The complex subclassification of pituitary adenomas is now recognised to reflect specific clinical features and genetic changes that predict targeted treatments for patients with pituitary disorders.     

The immunohistochemistry of pituitary adenomas

A great deal of progress has been achieved in recent years in the field of immunohistochemistry of pituitary adenoma. Continued use of more new antisera and monoclonal antibodies against numerous hormones in the adenohypophysis have resulted in new approaches to classification of pituitary adenoma. However, new problems have been discovered, on the other hand, by large-scale studies in recent years. The great number of multihormonal pituitary adenomas and possible change of the immunohistochemically detectable hormone status in cases of recurrent tumours have particularly re-emphasised the need for new thinking about patterns of classification. It would appear somewhat problematic, in this context, to uncritically accept terms, such as ACTH cellular adenoma or GH cellular adenoma. Reference is also made to the distribution pattern of cell and tissue markers in pituitary adenomas. The paper is based on thorough literature screening as well as on experience obtained by the authors from 450 cases of pituitary adenoma of which 260 had been analysed by immunohistochemistry, 131 by morphometry, and 80 by electron microscopy.     

MicroRNAs-based network: a novel therapeutic agent in pituitary adenoma

Pituitary adenomas are common benign intracranial neoplasms representing about 10-25% of all intracranial neoplasm. Significant morbidity can occur along with pituitary adenomas due to hormonal dysfunction and mass effects. The pathogenesis of pituitary adenoma is unclear, however, etiologic factors include genetic events, hormonal stimulation, and growth factors [1], all of which promote cell proliferation and transformation in the tumor. However, genetic events play the most important role in tumorigenesis. MicroRNAs (miRNAs), a class of non-coding RNAs, not only have function in pituitary cell proliferation and apoptosis but also in neoplastic transformation. It has been shown that miRNAs are differentially expressed in pituitary adenoma when compared with the normal pituitary gland; moreover, miRNAs have been identified as a predictive signature of pituitary adenoma and can be used to predict the histotype. The expression of miRNAs can be used not only to differentiate microadenomas from macroadenomas, but to also distinguish samples of treated patients from samples of non-treated patients. Therefore, we hypothesized that a miRNA-based network may be involved in pituitary tumorigenesis and it can potentially serve as useful diagnostic markers to improve the classification of pituitary adenomas. Here, we reviewed the therapeutic potential that different types of miRNAs may play in tumorigenesis. Moreover, miRNAs may emerge as potential therapeutic targets. We speculated the mechanism of miR-21 is involved in tumorigenesis, leading to improvements in therapies and prevention of metastasis.     

Hormone production andc-myc protein labeling in plurihormonal pituitary adenomas

The biological activity of plurihormonal pituitary adenomas was compared with that of tumors producing only one hormone by evaluating the percentage ofc- myc protein-labeled cells and ultrastructural characteristics. Twenty-five pituitary adenomas producing 3 or more hormones and 14 adenomas producing only I hormone were studied. Tissue sections were stained immunohistochemically using antibodies for pituitary hormones andc- myc protein, and they were examined by electronmicroscopy. DNA extracted from ethanol-fixed, paraffinembedded tissue was analyzed for p53 mutations by polymerase chain reaction and singlestrand conformation polymorphism analysis. The percentage ofc- myc protein-labeled cells in adenomas producing 4 or 5 pituitary hormones was significantly higher (p < 0.01 ) than in those producing 3 or 1 hormones. There were no p53 mutations in plurihormonal adenomas. Pituitary adenomas producing 4 or 5 pituitary hormones demonstrate biological aggressiveness; therefore, multihormone production reflects aggressive capacity rather than degree of differentiation.     

Hormone markers in pituitary adenomas: changes within last decade resulting from improved method

The significance of polyclonal antibodies for demonstration of hormone expression in pituitary adenomas was compared with the significance of monoclonal antibodies (MAbs). Adenomas were classified by light microscopic structures in paraffin-and epon-embedded sections, by immunostaining for all pituitary hormones, and in part by electron microscopy. In the first series, 166 adenomas were studied with polyclonal antibodies during 1990 and 1991. In the second series, 572 adenomas were immunostained with MAbs during 1999 and 2000. In the first series, a very typical type-specific hormone expression was demonstrable in 63.2% of adenomas in contrast to 91.1% in the second series. Consequently, in the first series 36.7% of adenomas expressed additional or atypical hormones in >10% of adenoma cells, whereas in the second series only 8.9% contained additional hormones. We conclude that MAbs show a clearer cutoff in immunostainings of pituitary hormones and there fore should be preferred for adenoma classification. Presented as a lecture on 8^th International Pituitary Pathology Meeting in Delphi (Greece), October 8, 2001.